Medium Bombers Chapter II

The Uneasy Years

On 11 November 1918, a Great War ended as the guns on the Western front and in other zones of conflict fell silent. But there was no general outbreak of peace. In Russia, a Civil War already raged that would last longer than the Great War and claim millions of casualties. That war would also spill over into a conflict between a re-established Poland and the USSR. Germany entered a period of political turmoil characterised by violence and the existence of small independent armed forces known as Freikorps. In China too peace remained out of reach, as different parties and warlords competed for power in a struggle that would not (temporarily) end until 1928. Large parts of Africa and Asia were being divided as the spoils of war between the victors of WWI, mainly Britain and France, who often enough needed to impose their power grab by force of arms. In French Morocco, combat operations would continue until 1934.

In a world that might be tired of war, but was by no means at peace, the air forces that had grown very rapidly in the crucible of WWI had to consider their future. Visionaries wrote books in which they declared that strategic air power would decide future wars. The practical need of the moment was often one for aircraft that were suited for low-intensity conflicts, affordable and safe to operate. To some degree, this set the stage for a division between “heavy” bombers designed to strike at an enemy’s industrial centres and population in a major war, and the “light” or “medium” types more adjusted to current missions and budgets. Often enough the needs were adequately met by cheaper single-engined aircraft, some light and fast, others relatively heavy.

Characteristic for the period was that multi-engined combat aircraft were often multi-role types, in the hope of achieving economies and getting the most out of a limited force. Britain, Italy and Germany invested in aircraft that can be defined as bomber-transports, a compromise that usually limited their usefulness as bombers, but provided enduring value as trainers and transports. France opted for “multi-seat combat aircraft” or “bomber-combat-reconnaissance” (BCR) types that were expected to fulfil multiple combat roles. The USA also experimented with multi-role designs, but typically aimed for repurposing a single airframe by installing different equipment for different missions.


In the early 1920s the existence of the Royal Air Force as an independent force was repeatedly scrutinised. Especially the Admiralty sought to win back control of aircraft for naval missions (with hindsight, not without good reason). The case for the RAF was built primarily on the requirements for Home Defence, the defence of the far-flung parts of the British Empire against major wars, and the usefulness of Air Control.

Air Control

Air Control, as first put into practise in Somaliland in the 1920s, was the use of air power in what now would be called counter insurgency warfare. Air attack proved its value for the British in Somaliland as it enabled the quick and cheap defeat of the forces of Mohammed Abdullah Hassan, where otherwise a lager and more expensive commitment of ground troops would have been necessary. Soon air strikes against rebellious groups and villages became a standard tool of controlling the British Empire. Arguably, these colonial operations laid the foundations for a form of air warfare that has remained hidden in plain sight from 1920 right until this day, as (mostly) Western powers have continued to use their superior air forces as a relatively cheap and risk-free means to exercise their power in the global south, combat irregular forces, or launch punitive strikes.

Given the absence of any opposition in the air and the fact that the targets were often mere camps or villages, Air Control did not require sophisticated or large aircraft, but the hostility of both terrain and people required reliable ones. Single-engined biplanes such as the D.H.9 and its successors were sufficient in most respects. But the British, and later the French and Italians, soon discovered a requirement for a “colonial” bomber-transport type. Typically this was multi-engined, able to operate from rough fields, and able to fly both bombardment and transport missions, with transport being the primary role.

A first example of the type was the Vickers Vernon, which entered service in 1921 as a military adaptation to the “Vimy Commercial”, itself an airliner development of the Vimy bomber. The Vernon had a primary transport role for which its bulky, rotund fuselage was well suited, but could also carry bombs. The successor to the Vernon was the Victoria, a development of the Virginia bomber with a similar bulky fuselage. The Valentia, a development of the Victoria with Bristol Pegasus radial engines, was able to carry bomb racks and flew bombing missions as late as 1940, by which date it was thoroughly antiquated.

Vickers Valentia (WikiMedia)

Specification C.26/31 issued in April 1932 sought a replacement that wasn’t so firmly rooted in WWI technology, a twin-engined bomber-transport suitable for operations in tropical climates and desert conditions. Primarily a transport with a capacity of 24 troops or 3000 lb (1360 kg) of cargo, it also needed the ability to drop eight 250 lb general-purpose bombs. The winner of this competition was the Bristol Bombay. The Bombay was in many ways resolutely modern, with all-metal stressed-skin construction, a streamlined fuselage, and neatly cowled Bristol Pegasus radial engines. But when delivered in 1939, its fixed landing gear and high-wing configuration were outdated for a frontline bomber, as well as its 309 km/h (192 mph) maximum speed. For defensive armament the Bombay featured hydraulically powered nose and gun turrets, type Bristol B.II and B.III, which was another sign of modernity, but each held only a single .303 Vickers K machine gun. Only 50 would be built, destined for use in the Middle East. Later No 216 Squadron used them for raids on Italian-held Benghazi, but it was evident that combat roles required more modern bombers.

Bristol Bombay. This is the prototype Bristol 130, as prepared to be shown to the public at RAF Hendon in 1936. (Bristol Aeroplane Co Ltd. via WikiMedia)

One of the competitors for C.26/31 was the Handley Page H.P.51. While it lost to the Bombay for this requirement, the RAF was in need of an interim heavy bomber until modern equipment came along, and that resulted in an order for 100 of the modified HP.54 Harrow bomber-transports to specification 29/35. Because of its very similar configuration (high wing, fixed undercarriage with large wheel spats, and Bristol Pegasus radial engines) the Harrow is easy to confuse with the Bombay, but it was a technically more conservative design and was delivered sooner, in 1937. Phased out as bombers by 1939, they served as transports, successfully but with decreasing numbers, until May 1945.

Amazingly enough, specification C.26/31 also gave the impetus behind a third bomber project. Armstrong Whitworth proposed its design A.W.23, which was built and flown in prototype form. It did not enter service, but became the ancestor of the Whitley.

Home “Defence”

While Air Control helped to make the case for the survival of the RAF, the priority of the Air Staff in these years was Home Defence, with the possibility of a major war in the Far East (due to the emergence of Japan as a major power) a secondary consideration. In 1923, as Germany was at least temporarily eliminated as a threat, the potential enemy for planning purposes was France. A conflict with France was by no means unimaginable, because of recurring friction between London and Paris over their conflicting colonial ambitions. With the goal of having parity with France, a 52-squadron force was designed, though the full implementation of this would be much delayed. The principle behind this scheme, as a 1923 report stated, was that “offensive action by aircraft in the enemy’s country is the best form of defence.” It was conceded that some defensive measures would be necessary, but the force structure was heavily biased towards bombers, which would equip 35 squadrons out of the total 52. As the targets were thought to be in North-Eastern France or near Paris, great range was not needed.

No consensus existed on what type of bomber best suited the RAF, however. In 1923 it was agreed to equip 12 squadrons with night bombers and 12 with day bombers, with the remainder to be decided later. By 1928 the Air Staff distinguished four bomber types: “high performance day bomber”, “medium performance day bomber”, “night bomber”, and “giant night bomber”. (In addition, there was a “general purpose bomber reconnaissance for overseas.”)

Of these, the high-performance day type was what later be considered a fast light bomber: In 1928 the Fairey Fox had entered service and the Hawker Hart was in development. These were beautifully streamlined, single-engined biplanes that were fast enough to elude interception by contemporary fighters. The “giant night bomber” was perhaps best represented by the Vickers Virginia, a large twin-engined biplane that was a fairly straightforward development of the Vickers Vimy of the end of WWI. In 1928 the Virginia was about to be replaced… by the Virginia, although in an updated version that replaced wood with metal throughout. Although both were biplanes, the Virginia and the Hart were poles apart in appearance and performance, and it is hard to believe that both were in production at the same time.

In between them, what concerns us here, were not one but two medium bomber types: A day and a night bomber. The RAF wanted to be able to hit targets both during day and night, but in 1928 the perception was that this required two distinct types. At the time, instrument flying was in its infancy: Jimmy Doolittle would make his famous instruments-only flight in 1929, and vital equipment such as the artificial horizon was still in development. Radio navigation tools suitable for navigating over enemy territory would only be developed during WWII. Airfields were grass fields as a rule, and until wings were equipped with more effective flaps and slats, high flying speeds corresponded with high landing speeds. The night bomber thus was likely to be a slower aircraft, very stable, with a larger crew to make room for a dedicated navigator and a radio operator. The RAF hoped and expected that a future type would be “dual role”, that is a “day and night” bomber.

At the end of WWI, the development of new medium day bombers had encountered a major snag. A number of aircraft had been designed as potential replacements of the D.H.10 Amiens: The Airco D.H.11 Oxford, the Avro Manchester (first of its name), the Sopwith Cobham and Boulton & Paul Bourges. Unfortunately all were among the many aircraft designed to be powered by the ABC Dragonfly, a radial engine that was ordered in huge numbers in 1918 to equip a new generation of aircraft and turned out to be an engineering disaster of the first order. By ill fortune and lack of knowledge, the critical frequency of torsional vibration of the Dragonfly’s crankshaft matched its running rpm, giving the engine a strong tendency to shake itself apart, and that was not even its only serious flaw. By the time the contenders had been reworked to use better engines, the RAF had abandoned the plan to replace the Amiens. A medium night bomber, the Nieuport London, was also doomed by the ABC Dragonfly.

Single-engined bombers looked like a viable alternative in the 1920s, and the Air Ministry wrote specification 2/20 for a “single-engined, long-range, heavy day bomber.” In 1924 this lead to the service introduction, after prolonged development, of the Avro Aldershot, a substantial aircraft with a crew of four and an all-up weight of 4965 kg (10950 lb), powered by a single 650 hp Rolls-Royce Condor III engine. It could lift eight 250 lb bombs internally or four 520 lb bombs externally. But the top speed of 177 km/h at sea level (110 mph) was inferior to the D.H.9 and D.H.10 of years earlier, and perhaps in recognition of this the single squadron of Aldershot bombers flew mostly at night.

It was recognised that twin-engined bombers were likely to offer more operational flexibility and safety. John North at Boulton & Paul had not given up after the failure of the Bourges. In 1922 the Bolton flew with an all-metal construction and Napier Lion engines, followed in 1923 by the improved Bugle with Bristol Jupiter radials or Napier Lion in-line engines. In recognition of these efforts, in 1924 Boulton & Paul was issued Specification 9/24 for an “Twin-Engined Medium Range Day Bombing Landplane” which became the Sidestrand.

Second prototype of the Boulton & Paul Sidestrand. The fuselage looks rather rotund in this picture but was in fact slim and deep.

The Sidestrand looked more modern than the Bourges, Bolton and Bugle, thanks to a redesigned, well-streamlined metal fuselage. It initially lacked the agility that pilots had appreciated in the earlier designs, but this was restored after some development of the control surfaces, notably adding a large Flettner tab that trailed behind the rudder. It had a modest bomb load (four 250 lb or two 520 lb bombs) and with two 460 hp Jupiter VIIIF radials, a top speed of 224 km/h (139 mph) loaded. It entered service in 1929 with No 101 squadron, but the 18 aircraft completed were only sufficient to equip that single unit. The Sidestrand, after problems with handling and engines had been resolved, was liked well enough as an aircraft. However, the single squadron with its ability to deliver relatively light bomb loads over a short range did not impress.

The RAF still operated 25 squadrons of Hawker Hart bombers against one squadron of Sidestrands, and firmly believed the single-engined type to be more suitable. The Hart could carry the same bomb load over the same range. The main advantage of the Sidestrand over single-engined bombers was its superior defensive armament, with open positions in the nose and in a dorsal position, plus a ventral gun to defend the lower rear. A disadvantage of the larger aircraft was that it was harder to disassemble and ship, should there be a need to redeploy aircraft to the Far East.

The final iteration of the design was probably the most important one. Originally known as the Sidestrand Mk.V and projected as an upgrade replacing the Jupiter engines with 550hp Bristol Pegasus engines, it involved a substantial redesign. The Overstrand featured Bristol Pegasus engines with Townend ring cowlings, a hydraulically powered nose turret (a first on a British bomber), an enclosed cockpit, strengthened landing gear, a 50% increase in bomb load, improved control surfaces, and an autopilot. The Overstrand replaced the Sidestrand in No 101 squadron from 1935 onwards. As new monoplane bombers were clearly the future, the number of Overstrands remained limited to 24 aircraft plus three conversions from Sidestrands.

Boulton & Paul Overstrand

Thus, before it started re-equipping itself in the run-up to WWII, the RAF relied mainly on single-engined, fast day bombers and heavy night bombers (such as the Virginia, Hyderabad, Hinaidi and Heyford). The twin-engined medium bomber was a category represented by a small number of aircraft that equipped a single squadron. At its state of development, the advantages of the type appeared to be minimal.

It is often assumed that under the influence of Trenchard, the RAF’s policy between the World Wars was very bomber-centric and focused on a strategic bomber offensive. The actual reality of the equipment fell short of that, however. It would also have significant consequences that the RAF had no experience of major combat in these years, nor were British bombers (apart from the single-engined types) exported to nations that were engaged in combat.


France emerged from WWI victorious but exhausted. It had lost a huge number of men, years of fighting had devastated a substantial stretch of Northern France, and financial recovery was expected from German reparation payments. It nevertheless resolved to operate a substantial air force. The law of 1920 proposed an air force of 135 escadrilles, of which 24 would be day bomber and 12 night bomber units. There would be 27 fighter units and the remainder would be very numerous reconnaissance and observation units. French forces were still engaged in combat in newly (and not so newly) established colonies.

The bulk of the bomber squadrons were equipped with late-war types such as the Bréguet 14 and 16, and the Farman F.60 Goliath heavy bomber. Characteristic for French policy in these years was that relatively modest numbers of aircraft were ordered for production, but the state subsidised development of prototypes in the hope of encouraging manufacturers to remain technically current. In practice, technical developments were slow. By 1929 night bomber units were flying the Loiré et Olivier LeO 20 Bn4, a twin-engined four-seat biplane, while the large majority of day bomber units (15 escadrilles with 232 aircraft) were equipped with the Bréguet 19, a single-engined two-seat light bomber that was a successful replacement for the Bréguet 14. Both represented only modest technical advancement.

New equipment arrived in the form of the Amiot 122 Bp3, a large three-seat biplane, of all-metal construction with fabric covering. A wing span of 20.04 m, an empty equipped weight of 2677 kg, a loaded weight of 4300 kg, and a maximum take-off weight of 5300 kg made this a lot of aircraft for a single 785 hp Lorraine 18KD engine. It has been described as France’s last single-engined heavy bomber, and indeed it was nearly as heavy as the twin-engined LeO 20 night bomber. (It can be thought of as the French equivalent of the Avro Aldershot.) The production run of 80 aircraft was relatively successful however, and the type was finally withdrawn from service in 1936. Nominally, the Amiot 122 Bp3 equipped heavy bomber squadrons. The type could carry up to 1100 kg of bombs.

France did operate an aircraft that looked a but like a medium bomber and that was operated by bombardment units, but this Blériot 127 M4 was a “multiplace de protection“, an escort aircraft for bombers, with only a limited capacity to carry light bombs. A typical creation of the transitional period in aircraft design, the 127 was an angular monoplane, with two gunners sitting in elongated engine nacelles and another one in the nose. It pointed towards a specific French trend that would result in a number of multi-seat multi-role designs.

A new sense of urgency developed almost immediately after Adolf Hitler became chancellor of Germany on 30 January 1933. In March, the French Ministry decided to re-equip of the French air force with modern aircraft. This became the Plan I, finally approved in October as the political situation continued to worsen. Of the planned strength of 1010 aircraft in the first line, 350 would be bombers, of which 120 heavy bombers, 16 very heavy bombers, 34 aircraft dedicated to naval missions, and 180 medium bombers. Plan I amounted to a reduction in quantity in the interest of rebuilding quality, as the available financial means were so limited. In July 1934 the French parliament nevertheless voted to spend substantial extra funds on defence.

But the French aviation industry was unable to deliver aircraft in the required numbers, and the quality was cause of concern. In the medium bomber class, France produced aircraft such as the Amiot 143, Potez 540, and Bloch 200, relatively successful but clearly outdated in comparison with the Douglas DC-2 or Martin B-10. As for production rates, although the workforce more than doubled between 1933 and 1935, deliveries on 1 January 1935 had only been 231 aircraft instead of the hoped for 458. The shortfall on the production goals of Plan I would finally be 24%.

Despite the shortfall in production, after the outbreak of the Spanish Civil War France would supply aircraft to the Republican government. The Popular Front government of Léon Blum was ideologically sympathetic to the Republicans, and besides the strategic logic was obvious: A Nationalist victory would result in France being surrounded by Fascist regimes on three sides. However, because of conservative opposition to the arms deliveries and British pressure, few aircraft were sent and these were flown over unarmed. These included 18 Potez 54 bombers, seven Potez 542 and one experimental Potez 544, and four Bloch 210. None of them would be very successful.

The Potez 540 was available in reasonable numbers because in the 1920s and 1930s, Potez had built a large, modern factory at Méaulte in northern France. Unfortunately the aircraft itself was already obsolescent when it entered service. Originally described as a four-seat “multiplace de combat” or M4 type, it was expected to be a multi-role aircraft serving both the needs of tactical recognise and bombing. Characteristic was a roomy fuselage with a comfortable position for the aircraft’s commander in the front, though the modest bomb load had to be carried externally!

Potez 540 with Hispano-Suiza 12X engines. (Wikimedia)

A high-wing monoplane, it featured the typical slab-sided fuselage and nose gun turret of inter-war French bombers. It also had a dorsal turret and a ventral turret, making it a well-armed design for the period, especially as the French were early adopters of powered turrets on bombers. The ventral turret, for example, had an electric drive for rotation, though the gunner controlled the elevation of the 7.5mm Darne or MAC1934 gun manually. As a forlorn concession to aerodynamics, the section of the ventral turret that protected the gunner’s legs could be folded upwards when not in use. The two engines were in nacelles under the wing that also housed the retractable main landing gear. The Potez 54 was robust and easy to fly, and built with several engines: 147 Potez 540 with Hispano 12X engines; 59 Potez 542 with Lorraine 12H engines; and 12 Potez 543 with Gnome-Rhône 14K radials. Although the type featured heavily in the rearmament plans of the early 1933s, its poor performance caused it to be relegated to second-line duties before the outbreak of WWII.

A similar fate befell the Bloch MB.200, which was similarly angular and similarly armed, though it had fixed landing gear and two Gnôme-Rhône 14Kirs radial engines attached to the wing’s leading edge. Designed as a 4-seat bomber, it was structurally typical of the advanced designs of the early 1930s, with corrugated metal skinning of the wings and fuselage. This generation of aircraft quickly became obsolete, but the French built nevertheless 208 of these aircraft (the majority were built by Potez) and 124 more of the type were built under license by Aero in Czechoslovakia. The MB.200 could carry more bombs than the Potez 540 (1200 instead of 900 kg), but it was even slower and had a shorter range.

Bloch MB.200 (Wikimedia)

Bloch could do better and demonstrated this with the Bloch 210, an aircraft that showed its lineage from the 200: Still angular, still with corrugated skinning, but a low-wing aircraft with retractable landing gear. Powered by 950 hp Gnôme-Rhône 14N engines in its final form, the MB.210 was a five-seat bomber with better performance than the 200, able to lift 1600kg of bombs and operate over a longer (though still inadequate) range. The 210 was still in active service in May 1940, in the “Alps Operational Area” in the South-East, away from the front with Germany in the North.

That left the Amiot 143M as the older bomber that was, regrettably, still in service in May 1940 is significant numbers. Like its competitors, the 143M was compromised by the desire to have ample observation windows in what amounted to a lower deck, to the extent that there was a position for the second pilot below that of the first pilot. The navigator and the bombardier also sat in this ample lower deck, while the nose gunner and dorsal gunner shared the upper deck with the pilot. It was a lot of drag for the two 780hp Gnôme-Rhône 14Kdrs engines to pull. Up to 1800 kg of bombs could be carried internally and under the wings. The all-metal 143 looked antiquated but had been a modern design when it first flew; by 1940 it would be seriously out of its place.

Amiot 143M (Jane’s All the World Aircraft 1938)

On 25 August 1936 the French parliament voted into law a new Five-Year Plan for the air force. This envisaged the production of 1290 aircraft between 1937 and 1941. Significantly, only 53 of these were planned to be heavy bombers, while 388 were medium bombers and another 279 aircraft would be BCR multi-role or A3 reconnaissance types. The medium bomber types ordered as part of the 1937 batch would be the Bloch 210, and the BCR types the Amiot 144 (a modestly improved, and failed, version of the 143) and the Bloch 131 or 132 (only the 131 would be produced). This plan proved to be short-lived as only a few weeks later, the government considered further reinforcements and the air force proposed to increase its first-line strength by 50%. The reinforcement of the bomber force was an important element of this Plan II. It still envisaged that the first-line force would have 492 medium bombers and 72 heavy bombers. The additional expense was approved in February 1937.

The Bloch 131 represented a modest step in the right direction. Its genesis lay in a 1933 specification for another multi-role BCR (“Bombardment, Combat, Renseignement”) type, to which the air force ordered the Bloch 130 in October 1935. However, when the air force wisely dropped the fighter part of this requirement and desired a reconnaissance-bomber, Bloch radically reworked his initial 130 into the much more streamlined 131. The last traces of corrugated skinning, still present on the prototype, were soon gone. Gradual refinement result in a rather elegant aircraft with a dorsal turret, a 7.5mm MAC 1934 in a nose mount, and a third MAC 1934 in a ventral gondola, which was found to be a better solution than a ventral turret. The type was ordered in April 1937 and deliveries began in June 1938. In the end, of 142 aircraft ordered, the air force would receive 138. The Bloch 132 with Hispano-Suiza 14A engines was not built, so the only production model was the 131 with Gnôme-Rhône 14N engines and a top speed of 385km/h at 4000m. On the other hand, its radius of action with an 800kg bomb load was only 420km. More a light bomber than a medium bomber, the 131 too was obsolescent went the war broke out.

Bloch MB.131 (Wikimedia)

By January 1938, l’Armée de l’Air operated 195 Bloch 200, 119 Amiot 143, 244 Potez 540 and 542, and 144 Bloch 210 medium bombers, and only 20 Farman 222 heavy bombers. But none of these aircraft could be considered modern. Their angular lines visually advertised that the French air force was a generation behind the competition. And if quantity has a quality of its own, then the French intelligence services estimated that the German Luftwaffe had about 1600 bombers in service.

French policy in the final stretch before WWII would be determined by the Plan V of 1938. This envisaged another large increase in air force strength towards 4,739 aircraft in total, with 2,617 in the first line. But Plan V did diversify the bomber force in favour of lighter aircraft. The heavy bomber force was reduced to 40 aircraft (24 in the first line), and medium bombers to 450 (264), but it added 470 light bombers, 184 dive bombers, and 122 ground attack aircraft. On the positive side for the medium bomber force, new aircraft types were appearing, such as the Lioré et Olivier LeO 451 and Amiot 350.

In his valuable book The Rise and Fall of the French Air Force (French Air Operations and Strategy 1900-1940), Greg Baughen argues that the French air force policy was unduly influenced by the heavy bomber doctrine and that this contributed to its downfall. As the numbers above indicate, that wasn’t really the case. French policy finally envisaged the production of a symbolic number of obsolescent Farman 222 heavy bombers. It real doctrinal weakness was the production of large numbers of reconnaissance and army co-operation types. While striving for effective cooperation between the services on the battlefield was an excellent idea, the specifications around which these aircraft were built were badly mistaken, and they mostly were a waste of resources. (British development of the Westland Lysander indicates that the RAF was also on the wrong track.) Plan V included a belated shift towards more effective dive bombers and ground attack aircraft.


The pivotal event of the interwar years in Italy was the appointment of Benito Mussolini as prime minister on 30 October 1922. It was not so much a coup, as the pathetic collapse of a conservative establishment after years of internal unrest and violence by political gangs. One of the leaders of the Fascist gangs was Italo Balbo, a charismatic thug from Ferrara, who in 1926 became Secretary of State for Air. Within a few years he would be a general and Air Minister. In 1930 and 1931 Balbo lead formations of Savoia-Marchetti seaplanes on transatlantic flights. For a Fascist regime that embraced modernity, militarism and the symbols of state power, the propaganda value of its aviation industry and air power was obvious. But the combination of poor industrial management and lack of resources would make it hard to keep up appearances.

The workhorse of the air force in these years was the prolific Caproni Ca.73 family, in different variants and at different times also known as the Ca.82, Ca.87, and Ca.74. What they all had in common was a low-slung fuselage, biplane wings of which the upper wing was considerably smaller than the lower wing, and two engines installed between the wings, one with a pusher propeller and one as a tractor, but not always in the same nacelle. The original Ca.73 had French Lorraine-Dietrich 12DB engines, but the Lorraine-Dietrich 12EB, Isotta-Fraschini Asso, and Bristol Jupiter powered other variants. As Caproni used the same basic concept to develop the larger, four-engined Ca.79 and the huge six-engined Ca.90, it is fair to consider the Ca.73/Ca.74 series medium bombers. The series participated in some of the first large paratroop exercises (having room for four parachutists) and was used on colonial warfare in Libya.

Caproni Ca.73ter bomber, a variant powered by two 500hp Isotta-Fraschini Asso engines. (Wikimedia)

Though a successful aircraft, the Ca.73 was warmed-up WWI technology, with a top speed of 174km/h and a range of 500km. In entered service in 1925, which happened to be the same year in which the Fokker VII/3m and Ford Trimotor appeared, and the Tupolev TB-1 flew. All aircraft that made the Ca.73 look like an antique.

Its successor in colonial operations was the Ca.101, a three-engined development of the single-engined Ca.97. The Ca.101 was a high-wing monoplane with fixed landing gear, not dissimilar in layout and appearance to a Fokker VII/3m or Ford Trimotor. It made its first flight in 1929. It was equipped with various moderately powerful radial engines; the Ca.102 was a twin-engined variant. The Ca.101 proved its worth when Italy invaded Abyssinia in 1935. In turn it was succeeded, from 1935 onwards, by the Ca.133, another three-engined high-wing monoplane. This retained the traditional mixed construction, but was aerodynamically somewhat cleaned up, and three 460hp Piaggio Stella P.VII radial engines gave it a top speed of 230 km/h. The Ca.133 was most useful as a transport (the Ca.133T) and in that capacity it would serve throughout the Second World War, becoming the Italian equivalent of the Ju 52/3m or the Harrow. In its own way it was a very successful type, rugged and versatile, with a production run of 417. But in the 1930s, such designs these were not viable combat aircraft in a European context. After many years in which Caproni had dominated the market for large aircraft, it was now struggling to be competitive.

Caproni Ca.133T

In 1934, Savoia Marchetti produced a successful airliner in the S.73. By adopting a low-set cantilever wing and omitting most struts and wires (though still retaining fixed landing gear), the S.73 achieved good aerodynamic performance, even if structurally it was conventional, with a welded steel tube fuselage and wooden wings. A military derivative was the S.M.81, which entered service in late 1935. This was built with various Italian and French radial engines, one of the more common ones being the Piaggio P.IX RC40 of 680hp. It is credited with a top speed of 340km/h, thus handsomely outpacing the Ca.133. The defence of the S.M.81 was entrusted to gunners in dorsal and ventral turrets with two rifle-calibre Breda-SAFAT machine guns, powered in rotation, but manual in elevation. Two additional machine guns were installed in lateral windows. The S.M.81 did see combat in Spain, and although obsolescent as a bomber by the outbreak of WWII, the rugged and reliable aircraft would serve on all fronts as a transport, and a few even remained in service post-war.

Savoia Marchetti S.M.81 (Wikimedia)

Thus in the late 1930s Italy possessed a modern medium bomber that was constructed in large numbers (535 built) and had good performance for its day. When it saw combat in Spain it was more advanced than the German Ju 52/3m, but it belonged to an older generation than the Heinkel He 111 and the Tupolev SB. The problem of its frontline aircraft being half a generation behind the competition would continue to handicap Italy.


There is a tendency to idealise the Weimar Republic as a liberal, democratic experiment in Germany’s history, highlighting the contrast with the military dictatorship that ruled Germany at the end of WWI, and the Nazi dictatorship that brought about WWII. And that is in some ways fair enough – but it doesn’t mean that German politicians, soldiers and industrialists all intended to abide by the terms of the Versailles treaty, which clearly banned aircraft capable of offensive operations.

Dornier, for example, continued to build large flying boats, based on its WWI experience, which resulted in the highly successful Dornier Wal series. For Japan it derived from this design the land-based Do N, which from 1928 onwards was built in Japan by Kawasaki as the Type 87 Heavy Bomber. In 1930 a land-based derivative of the Dornier Super-Wal flying boat became the Do P, a four-engined bomber that was tested in Switzerland. In 1931 followed the Do Y, a three-engined bomber of which three were delivered to Yugoslavia, powered by Gnôme-Rhône 9Kers radials, French versions of the Bristol Jupiter. A year later the Swiss branch developed the Do F, in theory a transport and mail aircraft, but in reality designed as a bomber. By Germany the Do F was operated in small numbers in nominal airline service, but the main purpose of these operations was to give future bomber crews experience in long-distance flying and night operations.

The military equipment of these aircraft was held in storage, but it cannot have fooled everyone as the Do F looked very much like a medium bomber, even if it was not a particularly advanced design for 1932. It had a metal structure, a high-set cantilever wing with a curved leading edge, two 550 hp Siemens Sh 22 (another version of the Jupiter) engines installed on the wing leading edges, an open cockpit and somewhat awkward, stalky, unreliable retractable landing gear legs. In 1933 the type was promptly converted in the Do 11 bomber, adding machine gun positions in the nose, dorsal and ventral positions. Though 122 of these bombers were completed, it was not a successful type. The Do 11D featured shortened wing tips and modified tail surfaces to improve handling. The Do 13 of 1933 was a development of the Do 11 that omitted the troublesome retraction of the landing gear, had slotted flaps and ailerons, and was powered in production form by BMW VI liquid-cooled engines. But only four Do 13s were delivered.

Dornier Do 23 (Wikimedia)

Further improvements, including structural strengthening, resulted in the Do 23 of 1934. Two 735 hp BMW VI U engines powered a high-wing bomber with an angular fuselage, fixed landing gear, slotted flaps, and a crew of four in open cockpits. It could carry a bomb load of 1000kg internally. Of these 282 would be delivered to the new Luftwaffe. It was very much a transitional type, as the fast Do 17 was then already under development. The Do 11, 13 and 23 family made the contemporary French bombers look almost elegant. But they were useful in allowing the new air force to equip its first bomber units, and indeed starred in propaganda movies about Germany’s rebuilt airforce.

Significant competition came from Junkers. The Ju 52/3m was primarily a transport aircraft, and would serve in that role through WWII and (in French service) even beyond it. In 1934 it was advanced enough to become a makeshift bomber-transport, the Ju 52/3m ges. It was capable of lifting six 100 kg bombs, while its defence was entrusted to a dorsal gunner and a gunner in a ventral “dustbin” turret, awkwardly fitted just after of the main landing gear legs and folded upwards when not in use. No doubt making the drag penalty bad when not deployed, and very bad when deployed. (The basic Ju 52/3m at least was slightly faster than the Do 23.) The bomber-transport version of the Junkers was nevertheless a bigger success than the Do 23, and at its peak the Ju 52/3m ges and the more powerful g3es equipped two-thirds of the bomber force. The type served with the Condor Legion in Spain and was also delivered to the Spanish Nationalist air force. But its days as a useful bomber were limited.

Thus Germany’s hastily rebuilt air force was, when it came to bomber strength, not all that impressive. The Junkers Ju 86, Heinkel He 111, Do 17, and Ju 88 would later equip a far more potent medium bomber force.


After the Revolution of 1917 and the end of the Great War, the former Russian Empire became the USSR, but the process involved a bloody civil war, the presence of foreign intervention forces, and an unsuccessful war with Poland. Furthermore, the USSR found itself ideologically opposed to most of its neighbours and diplomatically isolated. A notable exception to that isolation was the secret understanding with Germany, which among other things resulted in the production of Junkers aircraft in the USSR. One of them was the Junkers K 30C, alias the JuG-1 alias the R-42: A three-engined medium bomber based on the G 24 transport. A small number of these were built, but the Junkers design principles strongly influenced later Soviet designs.

Many of these designs did bear the name of Andrei Nikolayevich Tupolev, because in the 1920s and 1930s his was the principal design bureau, in which many talented Soviet engineers acquired experience before they were allowed to set up their own design bureaus. Tupolev’s first landmark contribution to bomber design was the ANT-4, with the official designation TB-1. The ANT-4 was a large twin-engine monoplane bomber, built out of duraluminum alloy with corrugated skinning of wings and fuselage. Though the prototype flew in 1925 with Napier Lion engines, production aircraft were powered by the 380 hp M-17, a Soviet version of the BMW VI. The 216 production aircraft were built between 1929 and 1932. The angular TB-1 had a crew of five or six, could carry a metric ton of bombs, and was defended by three open gunnery positions with a pair of DA machine guns each. (There was one nose position and two dorsal positions, but no ventral defence.) The TB-1 was not particularly fast, with a maximum speed of 196 km/h, which makes the retention of an open cockpit and open gunner positions more understandable. It had a range of 575 km with a full bomb load and 1000 km without bombs.

The TB-1 entered service as a heavy bomber, so it is debatable whether it can be considered a medium bomber. But Tupolev would follow up with the ANT-6, alias TB-3, a four-engined heavy bomber. The TB-3 owed a lot to the TB-1, but had a four M-17 or M-43 engines, a much longer wing span (39.5 m vs. 28.7 m), and approximately twice the loaded weight (10,967 kg vs. 4,520 kg). Deliveries of the TB-3 began in 1932, to a total of 819 aircraft. As the TB-3 was built by the same GAZ-22 factory that had built the TB-1, it supplanted it on the production lines. Tupolev would follow up with the even larger ANT-16 and ANT-20 designs.

The URSS-300, a record-breaking long-range version of the TB-1. (Wikimedia)

Incidentally, the first production TB-1 was completed as the unarmed URSS-300, named Strana Sovyetov (Land of the Soviets), and used for an eastward flight from Moscow to New York. The 21,242 km flight (with multiple stops) was finally completed by a second aircraft after the first was damaged. The link between Soviet bomber designs and long-range flights would continue with later aircraft, as in 1934 it was attempted to design a twin-engined long-range bomber on the basis of the single-engined ANT-25 record breaker. The ANT-25 had crowned its career with a flight over the north pole from Moscow to San Jacinto in California, 11,500 km non-stop. The ANT-37 (alias DB-2) retained the design of the ANT-25’s wings and tail, combined with two engines and a new fuselage. It did not enter production but two were used to set additional long-distance records.

Tupolev did not have the monopoly on bomber designs, however. In 1936 a most unusual shape climbed into the sky in the form of the Kalinin K-12, a tailless bomber. The K-12 featured a short fuselage with fore and aft gun turrets, each equipped with a single 7.62 mm ShKAS. Rudders were installed at the tips of the large, straight wing. Despite handling difficulties, a small series was ordered, but was cancelled in 1938. The more conventional K-13 was even less successful.

A far more significant aircraft was designed in response to a 1933 requirement for a fast bomber, under the SB designation, for skorostnoi bombardirovshchik, high-speed bomber. The concept was akin to that of the later Mosquito, a bomber fast enough to fly missions unescorted, although the SB would carry the traditional defensive guns in dorsal, ventral, and nose positions. The design goal was 330km/h at 4000m and the aircraft was to carry a 500kg bomb load. To achieve these goals, the design team opted for a semi-monocoque, smooth stressed-skin duralumin construction instead of the corrugated skinning of the earlier bombers, and initially also used flush riveting throughout. (Later, to simplify production, the use of flush riveting was restricted to critical areas.) The SB was also deliberately kept small. It had enclosed cockpits, retractable landing gear, and internal bomb stowage. The first prototype flew on 25 April 1934, powered at first by imported Wright SR-1820-F3 radial engines. But this aircraft was soon superseded by the second prototype, somewhat larger and heavier, and powered by imported 860 hp Hispano-Suiza 12Ybrs liquid-cooled engines. It flew on 7 October 1934 and had excellent performance, beating the specification by a 100 km/h margin.

This ANT-401 was far from trouble free, however, suffering from flutter problems that required incremental redesign of the control surfaces to cure. Even in its developed ANT-402 form, the type had numerous technical defects that affected maintenance and crew effectiveness, which lead to friction between the design team and the staff of the NII (the Scientific Research Institute of the air force). An issue that was never resolved was the inability of the navigator in the nose (also bombardier and gunner) to escape from a damaged aircraft, either in the air or after a wheels-up landing.

Production had already been started, however, and in 1935 the first series aircraft were delivered. The initial production version was powered by two 750hp M-100 engines, a Soviet version of the 12Y. It was larger and heavier than the prototypes, and not as well built. The maximum speed at 4000 m had decreased to 393 km/h. In 1936, the introduction of the M-100A, with an 860hp war emergency rating, restored it to 432km/h. However, modifications as a result of combat experience over Spain, including a new gun mount and the addition of armour plate to protect the pilot and navigator, added weight and reduced speed again to 412 km/h. The SB was not given self-sealing fuel tanks, despite an acquired reputation for easily burning when hit, and this weakness would persist into WW2.

The SB-2M-100A was very successful in Spain, flying its first combat mission against Tablada airfield near Sevilla on 28 October 1936. The SB was not particularly well defended, with dorsal and ventral mounts for a single ShKAS and two more machine guns in the nose, the latter having slots that permitted only vertical travel. The ventral gun was often discarded because it was too ineffective. But the speed of the SB was adequate protection as long as the main fighter on the Nationalist side was the Fiat CR.32 biplane, with a top speed in level flight of 360 km/h. The CR.32 could catch the SB only if it found itself in a favourable position for a diving attack. This changed in July 1937 when the Messerschmitt Bf 109B appeared over Spain. Though with a top speed of 465 km/h at 4000 m, even this fighter did not have a great margin over the SB. Attrition was nevertheless significant, and of 93 delivered, only 19 survived the war long enough to be capture by the Nationalists after their victory.

Incidentally, sheer disbelief that the USSR could produce an aircraft as advanced as the SB resulted in the type being referred to as a “Martin” aircraft in the press reports of the time. As late as August 1938 the Nationalists exhibited a captured example as “Martin Bomber”! That the SB had also been exported to China and Czechoslovakia apparently did not catch the attention.

Tupolev SB-2M-103. This is one of the aircraft captured by Finland and put into service. (Wikimedia) Especially from this angle, the Tupolev SB looks more modern than the contemporary Martin B-10.

A further improved series was the SB-2M-103, built with the 960hp M-103 in more aerodynamic cowlings and driving three-bladed instead of two-bladed propellers. This change increased maximum speed to 450 km/h at 4100 m. (It must be considered that high speeds were achieved at relatively low weight, and a prototype of the SB-2M-103 was tested to fly at 419 km/h at normal weight but only 378 km/h at maximum weight.) This version was also equipped with underwing external bomb racks to increase its lifting capacity to 1500 kg. It reached service units in the spring of 1938, in time to see combat with the Japanese during the border conflict over Manchuria. The type would also see combat in the Winter War with Finland of 1940, and the growing losses in this conflict indicated the SB’s obsolescence.

Finally, designer Arkhangelskii updated the aircraft with a new wing, and this version entered production in 1940 as the Ar-2, powered by two M-105R engines in very streamlined nacelles, as the coolant radiators had been moved into the wing’s leading edge. The Ar-2 was capable of 475 km/h at 4700m, further improved to 512 km/h by reducing its weight, but it was essentially outdated. The VVS ordered it after it had been modified to operate in a dive bomber role, but only 200 were built.

The SB in its various iterations was one of the most important bombers of the 1930s, built to a total of 6656 aircraft between 1936 and 1940. Perhaps even more importantly, it successfully served in combat in several wars, validating the concept of a twin-engined, relatively small but fast bomber. In Spain, the Nationalists and the Italian and German contingents that supported them at first relied on Ju 52/3m and S.M.81 bomber-transports, which were clearly outperformed by the SB. This prompted the Germans to send small numbers of the He 111B, Do 17E and Ju 86D, while the Italians sent the S.79. Thus, if the SB pioneered the role of the modern medium bomber, the response was quick.

And it had the misfortune of lasting into frontline service until 1942: At the time of the German invasion of the USSR in June 1941, the SB still equipped the bulk of the Soviet bomber force. Pitted against a new generated of fighters, the type was no longer able to escape by merit of its speed, as both the Bf 109 and Bf 110 were faster, and its defensive armament was weak. Losses were high, and the SB force steeply declined from 1428 aircraft in mid June to only 875 survivors by late July.


The designation system adopted by the USA in 1919 distinguished Day and Night Bombers, and divided the latter into short-range and long-range types: The designations were DB, NBS and NBL. A new system introduced in 1924 made a distinction between Bombers (B), Light Bombers (LB), and Heavy Bombers (HB). There was not formally such a thing as a medium bomber then, except by implication. Many aircraft would be re-numbered as the new system was introduced. For simplicity, we will mostly use the original designations here.

During WWI, American pilots had mostly flown aircraft of foreign design in combat, as the US industry was not ready to deliver competitive combat types. Work was ongoing to remedy this, of course, but the end of the war would cut short the production runs of many of these new aircraft. One of these was an aircraft that the US Army Air Services referred to as the GMB, or Glenn Martin Bomber, but is also commonly called the MB-1. (For Martin Bomber, not medium bomber.) Designed by Donald Douglas and put together under the supervision of Larry Bell, the MB-1 was first flown in August 1918. It was a twin-engined biplane of conventional configuration and equally conventional wood and fabric construction, compact and powered by two 400hp Liberty 12A engines. It was able to carry a 470 kg bomb load and performance was decent, 168 km/h at sea level. It was seen as an aircraft that could be built in different versions to accomplish different missions: A reconnaissance version with two cameras, a day bomber, a night bomber, and a “gun machine” with a 37-mm cannon that was known as the GMC. But because of the end of the war, large scale production orders were cancelled, and instead the Army took a mere handful of aircraft equipped to different standards, including one fitted out as transport. The Navy also took ten modified aircraft equipped as torpedo bombers. The MB-1 appeared doomed to be a vehicle for experimentation, a not unreasonable fate for the Army’s first home-grown bomber.

But from 1920 there was a major production development (130 aircraft) known as the NBS-1 or often as the MB-2. This was very similar in layout but different in detail: Simplified landing gear, an internal bomb bay, engine nacelles mounted on the lower wing that included the fuel tanks, and longer-span folding wings distinguished the MB-2. It was fitted with 420 hp Liberty 12 engines, and equipped to carry a much larger bomb load (815 to 1360kg) over short ranges. It was a bit slower than the MB-1, 159 km/h at sea level, but despite its NBS designation had a longer range, 900 km as opposed to 630 km. Significantly, the MB-2 was used in the 1921 exercises in which bombers attacked redundant ships and sank the former German battleship Ostfriesland. It was not easy for the American airmen to find a role for bombers, as there was no potential enemy nearby and the US public largely opposed sending forces overseas again to participate in foreign wars. The potential of bombers against enemy ships (a potential that, as far as level bombing was concerned, was much exaggerated) gave them an acceptable function to help defend budgets.

Martin MB-2 (Wikimedia)

Most MB-2s were not built by Martin, as under the contemporary procurement system the air force was free to ask other companies for bids to build the aircraft, which resulted in Martin, which had to amortise the development costs, being out-bid. This also resulted in Curtiss starting its own line of development of the MB-2, for which it was finally rewarded by an order for twelve Curtiss B-2 bombers. This was a development of the MB-2 with a steel instead of wooden construction, different wings, and gunner’s positions in the rear of each engine nacelle. Curtiss Conqueror engines replaced the old Liberty 12. With a roomier fuselage, this lineage would evolve into the Curtiss Condor airliner.

But the real successor of the MB-2 would be a family of bombers that was developed by Huff-Daland, or Keystone, as the company became known after a reorganisation. The original LB-1 of 1923 was a light bomber powered by a single 800 hp Packard 1A-2450, but was developed into a number of variants that were powered by two smaller engines. With the exception of the LB-5, which had two 420 hp Liberty 12 liquid-cooled engines, they were fitted with successive generations of the radial engines developed by Wright and Pratt & Whitney. Six different major engines and several minor variants resulted in a profusion of types and designations. Most were built only in small numbers (the largest run appears to have been 63 of the LB-10A alias B-3A type with 525hp R-1690-3 engines) and this must have become a maintenance and supply nightmare, though presumably it allowed the Air Service to trial the latest engines. The B-3A had a maximum speed of 183 km/h and could carry 1130 kg of bombs. The type had a respectable range, 1380 km, but compared to the old MB-2 the performance increase was modest. When production of the Keystone biplane bombers finally ended in 1932, they were relics of a past age.

Keystone B-3A (Wikimedia)

In an age in which the air service experimented with huge bombers such as the six-engined Witteman-Lewis XNBL-1 of 1923, or the heavily armoured Boeing GA-1 of 1921, the Martin and Keystone biplane bombers were evidently conservative. These were workhorse aircraft, and low risk designs. Other nations had independent air forces that developed bombers based on a strategic vision, but in the USA the Air Service was a branch of the US Army and air officers struggled to implement their ideas. For much of the 1920s, the US maintained only a single bomber group, the 2nd, in the continental USA, plus smaller detachments to outlying strategic areas such as the Panama Canal Zone, Hawaii, and the Philippines. Existing seems to have been their main purpose. At the court-martial of William “Billy” Mitchell in 1925, Carl Spaatz testified that only 22% of the service’s aircraft were fit for service. Fair or not, this reflected the frustration of the officers that would lead the Air Army Force in WWII, with its state in the 1920s.

In late 1929 the Air Corps circulated requirements for a new bomber. This time, the goal was to make a break with the past: When Martin submitted a design for another biplane bomber, it was promptly rejected and Martin received additional specifications and instructions instead. The development of a new aircraft was not without pitfalls however, and a series of transitional types emerged to be built in small numbers.

In June 1929, Fokker Aircraft Corporation — the American venture of the famous Dutch manufacturer — had received an order for two prototypes of the XO-27, alias Fokker 16, a new and innovative observation (i.e. tactical reconnaissance) aircraft. It featured the typical mixed construction of the Fokker monoplane transports. The wing was cantilever, of wooden construction, with a thick profile. The fuselage on the order hand was steel tube, partially covered with fabric, and partially with corrugated metal panels. It had retractable landing gear, although somewhat primitive, and two 600hp Curtiss V-1570-9 Conqueror liquid-cooled engines, largely buried in the wings. At the request of the air force, one of the prototypes was completed as the XB-8 bomber. Although testing was not entirely successful, in 1931 the air force ordered twelve aircraft under the designations YO-27 and Y1O-27 (identical, but under different contracts). These aircraft had a different engine installation and a redesigned fuselage. With 600 hp V-1570-29 engines, they had a top speed of 285 km/h. The small number of aircraft served with both observation and bomber units, and were generally appreciated, though the retractable landing gear was the cause of several accidents. The last survivors were struck from service in 1937. Technologically, the YO-27 was a contemporary of the Tupolev TB-1 (which entered production in 1929) but less advanced.

Fokker XB-8 (Wikimedia)

As mentioned, Douglas had produced a competitor in the form of its XB-7. This was very much prompted by the appearance of the XB-8 and XO-27. As in the case of Fokker, two prototypes were ordered from Douglas, one to be completed as the XO-35 observation aircraft and the other as the XB-7 bomber. They were followed by five Y1O-35 and seven Y1B-7 production aircraft. The B-7 featured a gull wing with strut bracing, Curtiss V-1570 Conqueror engines attached to the struts, and basic retractable landing gear. Despite looking a bit more old-fashioned than the YO-27, they had comparable performance characteristics. They had an incident-rich career, and the last were struck from service in 1938.

The 1929 Army specification resulted in additional technological advancement. Boeing built on its experience with the single-engined Boeing Monomail transport to create the model 214 and 215 bomber prototypes, powered by two Curtiss Conqueror in-line and top Pratt & Whitney Hornet radial engines, respectively. An all-metal semi-monocoque fuselage, a thick cantilever wing; and semi-tractable landing gear resulted in a rather sleek design – perhaps too sleek, as the five members of the crew (nose gunner, bombardier, navigator, pilot and dorsal gunner) were parcelled out over a long and narrow fuselage, unable to communicate. Still, after the first flight of the 215 in April 1931 the air service was sufficiently impressed to evaluate the 215 as the XB-901, later Y1B-9. It subsequently purchased the 214 as well to refit it with Hornet engines. Finally it ordered five improved Boeing 246, alias Y1B-9, bombers with 630 hp R-1860-11 Hornet radials. These aircraft, delivered between November 1931 and March 1933, would serve very briefly: By April 1935 the last was disposed off. This is a bit surprising as they were more advanced aircraft than the Y1O-27 and Y1B-7, and less troublesome. Though it was observed that at the 300 km/h that the Y1B-9 was capable of, the open cockpits were definitely uncomfortable. Plans to modify the aircraft with closed cockpits were never implemented. Perhaps there was a conscious decision to not allow the B-9 to evolve in a worthy competitor of the B-10.

Boeing Y1B-9 (Wikimedia)

As mentioned, Martin’s initial proposal of a biplane bomber had been rejected. In 1932 it came with the model 123, which had a deep but well-streamlined fuselage, cantilever wings, open cockpits, and Wright R-1820 radial engines with simple cowling rings. It was evaluated as the XB-907 but found to have deficiencies in handling and a high landing speed, so Martin had to re-work it into the XB-907A, with modified wings, improved landing gear, NACA engine cowlings, and a nose turret. With a top speed of 333 km/h, the XB-907 handsomely outperformed the competing designs including the B-9. Martin was rewarded with an order for 48 aircraft of model 139, divided between the B-10 with Wright R-1820 radials, the B-12 with Pratt & Whitney R-1690 radials, and one B-14 with the new Pratt & Whitney R-1830. These were fitted out with enclosed cockpits. In 1934 and 1935, Martin received an order for another 103 aircraft, all B-10Bs with 700 hp R-1820-33 engines. The new bomber soon equipped the all the bombardment groups. Its maximum speed of 345 km/h enabled the B-10B to evade most fighters, and in 1935 the B-10 was used for trials of the new Norden bomb sight. In the B-10 and B-12, the US Army Air Corps had a highly successful modern bomber, which was a contemporary of the Tupolev SB. The Martin would also be an export success, with the most important customer being the air force of the Dutch East Indies, the KNIL, which received 120 aircraft of the model 139 and improved model 166. Exported aircraft did see combat in the early years of WWII. In the US, the type was relegated to second line duties by the outbreak of WW2.

The B-7, B-8 and B-9 belonged to a generation of transitional aircraft that was built in small numbers before the introduction of the B-10, which was the culmination, at least in the short term, of the evolutionary progress. It is interesting that this happened in the USA while France, for example, remained stuck with bombers such as the M.B.200, Amiot 340 and Potez 540 that very much belonged to that transitional generation. This is often attributed to the emergence of the US civil aviation market, which combined a real demand for internal trans-continental transport with de facto government subsidies that rewarded investment in new technology such as the Boeing 247 and Douglas DC-2 airliners. Another major factor was NACA, where aerodynamic research made great strides in the 1920s, contributing a series of widely used aerofoil designs, as well as the NACA cowling design that revolutionised the efficiency of radial engine installations from the late 1920s onwards. While European powers did fund technology development and subsidise airlines, they were not in a position to do so to the same degree as the USA. NACA would influence aircraft design all over the world, but at a crucial time it gave the USA a decisive technological lead.

Martin B-10B (Wikimedia). The separation of the crew in different cockpits was a disadvantage of the type, and the later model 166 had a continuous greenhouse canopy.

Still, if the Martin B-10 and B-12 were advanced bombers for their time, they were not yet what the strategic bombing advocates in the air service really wanted. In 1935 the Air Corps wanted a bomber with double the range and twice the lifting capability. Martin’s decision to offer a development of the B-10 as the type 146 resulted in a failure as the Air Corpse preferred the offerings of Douglas and Boeing: The Douglas DB-1, which became the B-18 Bolo, and the Boeing model 299, which became the B-17 Flying Fortress. It is noteworthy that the B-17, which emerged as the most iconic “heavy” of WWII, participated in what was in many ways a medium bomber competition.

Because thinking about heavy bombers had evolved as well. In 1933 Materiel Division formulated a specification for a bomber with a range of 8000 km, if with a modest bomb load of 910 kg. In July 1935 the Air Corps ordered the Boeing XBLR-1, alias XB-15: A 32-ton giant with a wing span of 45m and four R-1830 engines. However, the underpowered XB-15 fell way short of its design goals. Even more ambitious was the Douglas XBLR-2, alias XB-19, which was not a direct competitor of the XB-15, but a leap beyond it: with a 64m wing span, a weight of 73 tons fully loaded, and powered by the new Wright R-3350 radials, the XB-19 was designed for a maximum range of 12400 km. It was later re-engined with Allison V-3420 in-line engines. Such designs dwarfed even the B-17 and pointed in the direction of the post-war B-36.

The Belgian poet Willem Elsschot famously wrote that between dream and deed stand practical objections (the deed in question being murder) and the same was true for the USAAC in the 1930s. The XB-15 and XB-19 were very impractical aircraft and even the B-17 was considered to be too expensive. The definitive pre-war medium bomber was thus the B-18, an odd-looking and unspectacular aircraft. It used the wings and tail of the DC-2 airliner with a new deeper fuselage, allowing for an internal bomb bay. Like other medium bomber of the era, it was defended by guns in the nose, a dorsal, and a ventral position. The 133 B-18s ordered in 1936 were followed by 217 B-18As with a different nose configuration and a more powerful version of the Wright R-1820 engine. The B-18 wasn’t any faster than the B-10, but it could carry nearly three ton of bombs. It was almost the aircraft that gave the medium bomber a bad name, as it equipped many US bomber units during the first years of WWII, but was clearly outdated by the time of Pearl Harbour. It was fairly quickly relegated to anti-submarine patrol duties. If its career was somewhat lacking in martial glory, it had the advantage of being around in numbers when he USA needed to start expanding its air force.

Douglas B-18A (Wikimedia)


In the 1920s, Japan was a rising power in Asia, eager to develop its aviation industry and air force. To accelerate technology development, the Japanese invited French and British technical missions, sent out engineering students to study abroad, and acquired licenses for foreign aircraft. This fostered the illusion that the Japanese were only capable of copying foreign designs, perhaps in part because French and British missions had been instructed to sell their products but keep the Japanese dependent, and they of course reported back that they had successfully done so. In reality the Japanese quickly became capable of independent, advanced designs. Though the relatively narrow industrial basis of Japan would continue to restrain production.

Under the Japanese constitution, the Imperial Army and Navy reported directly to the emperor, not to the civilian government, and as the imperial court was rarely in a strong enough position to exercise effective control over them, they developed into independent, uncontrollable and fiercely rival forces, each in control of their own military-industrial complex, and each pursuing their own strategy. In most nations, multi-engined bombers were the domain either of an independent air force, or (as in the USA) of the army air force. In Japan, the Army and Navy independently developed their own land-based multi-engined bomber forces. Of course, of the two the Navy had a bigger interest in attacking naval targets, which generally implied that in the trade-off between range and bomb load, it would opt for a longer range and a smaller bomb load, but with the ability to carry a torpedo as anti-ship weapon. Thus the Navy was more attracted to what conceptually be considered a medium bomber, but it adapted its designs to specific naval requirements.

One of the first twin-engined bomber projects of the Navy had the goal of designing a carrier-capable twin-engined bomber, under the designation of 7-Shi Twin-Engine Carrier Aircraft. But Mitsubishi’s biplane design, first flown in October 1932, finally entered service as the Type 93 Land-based Attack Aircraft, as it was not carrier capable. And because it was not very successful, only eleven were completed.

In the 1930s, with the expansion of the fleet and the construction of bases in the Pacific curtailed by treaties, the Navy developed an interest in long-range, land-based attack aircraft, backed strongly by then rear-admiral Yamamoto, the later commander of the Combined Fleet at the outbreak of WWII. This would result in two aircraft entering service in 1936, the Hiro G2H and the Mitsubishi G3M. (Nakajima developed a competing design, the LB-2, but the Navy did not accept it.)

Mitsubishi G3M2 bombers. The model 22 in the foreground is armed with a 20 mm cannon, Type 99-1, in a big dorsal turret, combined with 7.7 mm machine guns in a front dorsal turret and rear fuselage blisters. The model 21 in the background only had 7.7 mm (Type 92) machine guns in retractable dorsal and ventral turrets.

The G2H was the larger and the least successful of the two. Originating as the 7-Shi Special Attack Aircraft able to carry 2 tons of bombs and with a range of 2000 nautical miles (3700 km), the G2H had a wing span of 31.68 m and a loaded weight of 11,000 kg; its range fell short of requirements as it was only 1080 to 1557 nautical miles. The two 900 hp Type 94-1 engines were troublesome and handling characteristics left something to be desired. Only eight were completed, and had a brief service life as the Type 95 Land-Based Attack Aircraft.

The G3M was derived from the G1M design for a twin-engined 8-Shi Special Reconnaissance aircraft with an astonishing range target of 4000 nautical miles (7400 km) to scour the Pacific for the whereabouts of the US fleet. The G1M was a slender, attractive aircraft with twin tail fins and retractable landing gear, which flew on 7 May 1934. But the requirement had changed to define the 8-Shi Land-based Medium Attack Aircraft, and although the G1M performed well, the shift in operational requirements and the addition of a bomb load and defensive machine guns created some problems that were addressed by a redesign, the 9-Shi Land-based Medium Attack Aircraft, which became the G3M, with the official designation of Type 96 Land-Based Attack Aircraft. After a series of prototypes and pre-production aircraft, the bomber went into series production in 1936.

The G3M was thus a contemporary of the Douglas B-18 and the Tupolev SB, but it would remain in production until 1943, with a run of 1048 completed. It had a slim, tadpole-like fuselage and twin tail fins, though the streamlining was affected by the external carriage of an 800 kg torpedo or bombs. Production aircraft were powered by Mitsubishi Kinsei radials, in various versions developing between 1075 hp for the first models and 1300 hp for the later models. The G3M2 had an impressive range of 2635 nautical miles (4880 km) which was boosted further to 3363 nautical miles (6230 km) in the Nakajima-built G3M3 of 1941. But weak defensive armament (although no better or worse than other 1930s bombers) and the absence of any protection against enemy fire made the G3M a vulnerable target. The G3M saw extensive use against targets in China, after Japan attacked China in 1937, and its ability to hit targets deep into Chinese territory was limited by the serious losses suffered. The IJN to its credit understood the need for long-range escort fighters, which fed in the design of the A6M “Zero” fighter.

The G3M’s best known achievement was the sinking (together with the later G4M) of the British battleship HMS Prince of Wales and the battlecruiser HMS Repulse in late 1941, an event that has been interpreted as definitively marking the superiority of the aircraft over the battleship, as this was the first time that a battleship was sunk that while under steam and battle-ready. At this time about 250 remained in service, but the more modern G4M was already beginning to replace it in the first line. From 1943 onwards the aircraft survived in training, maritime patrol and transport roles, with a number being converted to transport configuration, the most thorough conversion being known as the L3Y.

The Army followed a rather different trajectory. It initially brough into service heavy bombers of foreign design. In 1927, the Type 87 Heavy bomber, built by Kawasaki but designed by Dornier in its factory in Switzerland. This was a large twin-engined aircraft that was in effect a land-based development of the highly successful Dornier War flying boats. Then, in 1931, the four-engined Mitsubishi Type 92, alias Ki-20, a derivative of the Junkers K 51, itself a bomber version of the Junkers G 38 airliner. This was obsolete by the time its development was completed. And in 1933, an enlarged derivative of the Junkers K 37 as the twin-engined Type 93 Heavy Bomber, alias Ki-1, to replace the Type 87. This aircraft did see combat over China, but it was not very successful and the army acquired Fiat B.R.20 bombers from Italy as a stop-gap solution until the arrival of modern heavy bombers of Japanese design, either from Nakajima (the Ki-19) or Mitsubishi (the Ki-21).

Ki-2 light bomber (WikiMedia)

The closest aircraft to a medium bomber in the Army’s inventory was the Type 93 Twin-Engined Light Bomber, alias Ki-2. Confusingly this had very similar origins as the Type 93 Heavy Bomber, both being derivates of the Junkers K 37. It looked broadly similar to the Ki-1 too, with angular outlines, a low-set monoplane wing, fixed landing gear, corrugated skinning, and a big nose turret. But whereas the Ki-1 was an enlarged derivate, the Ki-2 was optimised for performance and manoeuvrability, with stringent requirements on flying quality on one engine. Despite their similar appearance, it was a considerably smaller aircraft. The loaded weight of the Ki-2 (4550 kg) was less than the empty weight of the Ki-1 (4880 kg), and the light bomber carried 300 kg of light bombs with a maximum load of 500kg, while the heavy bomber was designed for 1000 kg with a maximum load of 1500 kg. Outwardly, the Ki-2 was most easily distinguished because it had Nakajima-built Jupiter radial engines instead of liquid-cooled engines. It did see combat over China and Manchuria, and was considered far more successful than the Ki-1. An updated version, the Ki-2-II of 1936, featured more powerful Ha-8 engines, retractable landing gear, an enclosed cockpit, and other aerodynamic improvements. This increased the type’s maximum speed from 256 km/h (159 mph) to 283 km/h (176 mph). By then it was clearly an outdated design, but the Ki-2 would have a useful life as a trainer after it was withdrawn from combat operations.

Smaller Nations

Twin-engined bombers were financially within reach of the smaller nations of Europe, depending on their level of ambition and experience, if perhaps only just so. A state with a small aviation industry was typically able to produce single-engined light bomber and reconnaissance types, such as the Polish P.Z.L 23 Karas. Multi-engined bombers were considerably more challenging. A shortcut could be taken by license-building foreign designs, and for example bomber derivatives of Fokker transport aircraft were produced in several countries, including Poland and Czechoslovakia. This was a good way to acquire experience in building larger aircraft, but a long-term desire to maintain independence or neutrality, or a shortage of foreign currency, might argue against it. And of course there was always also the hope that the export market would offset some of the costs of development.

Lack of experience could lead to odd designs. In Belgium, for example, the threat of German rearmament induced the military to produce a very detailed specification for a twin engined bomber and reconnaissance aircraft, concepyualized as a sesquiplane with fixed landing gear and a fuselage roomy enough to allow the crew to exchange positions in flight. Heavy armament was specified, to the extent that a fighter combat role was considered, and the ability of one of the other crew members to act as back-up pilot. But the resulting LACAB Gr8 and Stampe-Vertongen SV10 were predictably awkward and obsolescent designs. The SV10 fatally crashed on its third test flight, and the prototype Gr 8 lingered in a hangar until German bombs destroyed it in May 1940.

LACAB Gr 8 (WikiMedia)

Poland had its own unhappy story with the L.W.S. 4, nicknamed Zúbr (Bison). This type was developed in the 1930s from an earlier design for the P.Z.L. 27 airliner, which had already been rejected by the Polish airline Lot because it wanted the more modern Douglas and Lockheed aircraft. Not an auspicious beginning, and to boot the Zúbr was considered a transitional type from the start because the air force awaited the more advanced and very capable P.Z.L. 37 Los. When it emerged for its first flight in 1936, the P.Z.L. 30 Zúbr was a low-wing, mixed-construction monoplane with retractable landing gear and an ugly, boxy fuselage. On the good side, that fuselage was roomy and comfortable, the Zúbr could carry a ton of bombs, and the armament of twin machine guns in nose and dorsal positions, and one or two in a ventral position, was respectable. The type moved with its chief designer from P.Z.L. to L.W.S., which received a production contract for 15. But from there on development started to go badly wrong, as the type suffered from an understrength and overweight airframe, a problem made worse instead of better by installing more powerful Pegasus VIII engines in the L.W.S.4. Add to this an unreliable undercarriage, and the L.W.S.4 proved a useless aircraft.

Czechoslovakia had its own aviation industry, which in the interwar years built several bombers under license, including the Bloch MB.200, the Tupolev SB, and the more extensively modified Avia-Fokker F.IX, a derivative of the Fokker F.IX airliner. In 1938 the prototype emerged of an indigenous design, the Aero A.300 bomber, with pleasing lines and two Bristol Mercury radial engines. But the step-wise dismemberment of Czechoslovakia that followed the Munich agreement saw its industry diverted to the purposes of the invader.

Fokker T.V (WikiMedia)

In the Netherlands, Fokker developed the T.V, a multi-role type initially considered as a “cruiser fighter”. This designation reflected that the T.V was equipped with a Solothurn S18-350 20-mm anti-tank cannon in the nose. That was not an automatic weapon, but the responsible committee considered it ideal for long-range single-shot sniping at enemy aircraft, a concept that can be charitably described as overly optimistic. More realistically, the T.V was a bomber, and ordered in service as such. The T.V had clean, if boxy, lines, and a mixed construction. The biggest problem of the T.V was that at the time of the German attack in May 1940, only 15 had been delivered, of which only 2 were fully equipped. Its combat impact was therefore minimal.

No doubt the most promising bomber produced by a smaller state before WWII was the Polish P.Z.L. 37 Los. This was designed in 1934 as a fast light bomber, conceptually not dissimilar to the Soviet SB, and made its first flight in June 1936. Of all metal construction, with a streamlined fuselage that tapered to a slim tail, the compact Los had a crew of three and was powered by two PZL-built Bristol Pegasus radials. A bomb load of up to 2580 kg could be accommodated internally, in bays in the fuselage and the wing centre section. Because of the high expected performance, the air force decided against installing heavy armament, and instead defensive armament consisted of the then common standard of three rifle-calibre machine guns (7.7 mm KM Wz 37) in the nose, dorsal, and ventral positions. The radio operator was responsible for both the latter positions, while the navigator and bombardier operated the nose gun, commanded the aircraft, and had auxiliary flying controls; thus the crew no doubt had a high workload. The P.Z.L. 37 was reasonably fast: The 37/II Los B with 835hp Pegasus XX engines was credited with a a maximum speed of 445 km/h at 3400 m. It was also efficient, and had a range of 2600 km with a 1760 kg bomb load. There were some visual clues that indicated a less than fully modern design, such as the shape of the engine cowlings and the relatively large wing area (53.51 square meter) for its size and weight (8560 kg normal loaded). The first ten Los A production aircraft were built with a single fin and rudder, followed twenty Los A bis with twin tail fins, and finally an intended larger series of Los B in the definitive configuration. The type was also actively marketed for export.

P.Z.L. 37 Los bombers. The twin wheel on the main landing gear legs were and advanced feature for 1939, but the type suffered from landing gear failures when operating from soft grass airfields.

By the standard of 1939, the Los was a good bomber, aerodynamically advanced and with effective striking power. Plans were being made for a more powerful, larger P.49. Yet due to a protracted development history and opposition to the construction of bombers within the air force, the production of the P.Z.L. 37 for the Polish air force was cancelled in the spring of 1939! Instead of scaling up the production of the new bomber to meet a large demand, P.Z.L. found itself merely finishing the aircraft already under construction, while faced with shortages of engines and equipment. Thus when World War II broke out on 1 September 1939 with a German attack (soon followed by a Soviet attack) on Poland, only about 90 aircraft had been received in various states of equipment, with much less than half of these being in the front line strength. Combat losses and serviceability issues rapidly eroded this even more. At the end of a brief campaign, some were flown to Rumania, which briefly used them against the USSR before relegating the survivors to a training role. The history of Poland’s advanced bomber ended on a sour note, but not an uncommon one for such ambitious projects undertaken by Europe’s smaller powers.

A State of Transition

States with a powerful military all maintained bomber forces of some kind through the 1920s and 1930s, and they had some of the political impact that nuclear weapons would have after 1945. The theory that air power would be decisive in future wars, as proposed by Douhet, Seversky and others, might not in itself have been sufficient to exert such influence. But when combined with the known difficulty of intercepting bombers with the technical means of the time, a large over-estimation of the damage that bombing would do, and the widespread fear of the use of chemical weapons, this vision became apocalyptic. And as the threat of bombardment from the air influenced the decisions of politicians, they also sought the means to retaliate in kind, as a deterrent or as a means to grasp the initiative. A fleet of bombers was one of the hallmarks of a major power, and for the quickly re-arming Germany it had the significant advantage of being easier to produce than that other symbol of power, a fleet of battleships.

This state of mind was notoriously revealed by British prime minister Neville Chamberlain, who in September 1938 remarked “How horrible, fantastic, incredible it is that we should be digging trenches and trying on gas masks because of a quarrel in a far-way country between people of whom we know nothing.” Said trenches were not dug along a putative new Western Front, they were being dug in London’s parks. Harold Macmillan commented that the British people had been told that “the devastation of air attack would be beyond all imagination.” Air Staff estimates of 1936 might not have predicted devastation “beyond imagination”, but they did expect that 375 ton of bombs over six days would be enough to knock out London’s utilities, including all gas supply, half the electricity supply, and the Underground. Such elevated fears of enemy attack were not always symmetric with a belief in the striking power of the own force, and also in September 1938 the commander-in-chief of RAF Bomber Command advised against attempts to bomb Germany, which might well end in disastrous failure.

This was more realistic, because most air forces of the 1930s had modest capabilities. Modern bombers were not available in large numbers, and light bombers and multi-role aircraft often predominated as a means to economise on limited budgets. As the heavy bombers of WWII were on the drawing board or deployed in small numbers, the striking power of most air forces was provided by twin-engined medium bombers. Besides range and bomb load, there were other technological limitations, and for example some officers in the RAF were rightly pessimistic about its ability to find and hit targets in Germany, as exercises had demonstrated that navigation tools and bombsights were not up to the task.

The bombing of Guernica in April 1937 shocked the world, but these raids were conducted by only two dozen bombers and they dropped about 40 tons of bombs. That was a large effort by the standards of the Spanish civil war, and resulted in the death of hundreds of civilians, but events were about to become unimaginably worse.

Next: Medium Bombers Chapter III