External-spring type



The identity of the designer of the first successful indicator to isolate the spring from the effects of heat (usually by removing it from the piston chanber) is still disputed. Experiments undertaken by the Royal Navy in the 1840s, summarised in the book by Main & Brown, were assisted by an indicator made by Maudslay & Field. The engraving prepared for the book shows a suprisingly sophisticated instrument with a rotating drum and the spring carried externally on two standards.

Attempts have been made to date this design as early as 1822, but evidence is sketchy. If such an early date can be proven, however, there would be little doubt not only that the Maudslay & Field design was the first to mount the piston spring internally, but also that the rotating drum customarily associated with McNaught may have originated elsewhere.

Credit has also been given to the Scotsmen McKinnell and Buchanan, whose alteration of a basic Richards-type indicator was patented in Britain in November 1892. The spring, carried on top of two standards, acted on an extension of the piston rod; a rounded abutment on the otherwise conventional Richards amplifying linkage was actuated by a collar on the piston rod to move the trace point. It is assumed that a few of these indicators were made by Hannan & Buchanan of Glasgow, makers of Richards, improved Richards (including a small-scale 'high speed' pattern) and Thompson-type instruments prior to 1914.

Dobbie, 1898

A derivative of the basic Thompson linkage was the subject of a patent granted in Britain in 1898 to John Clark Dobbie, a 'Nautical Instrument Maker' of '45, Clyde Place, in the City of Glasgow'. The indicator was based on the proven McInnes pattern—which was little surprise, as Dobbie & Son traded next door to T.S. McInnes & Co. Ltd. The principal claims to novelty included a readily removable piston spring, exposed to the air, and a 'pencil-carrying arm curved at its forward end and having a downwardly projecting knee or branch'. Changes were also made to the drum. A patent granted in December 1899 then allowed claims for a lightweight pointer assembly made of aluminium 'or similar light metal or alloy' and a lightweight piston consisting of two hollowed discs. The lower disc was steam-tight, but the upper one was perforated to allow grit to pass through.

The Dobbie indicator was initially made by T.S. McInnes & Co. Ltd, but then by a new partnership of McInnes and Alexander Dobbie & Son, known as 'Dobbie McInnes Ltd'. Too few indicators have been examined to deduce exactly when the first instruments were made, though Engineering of June 1900 contained a brief review of the new external spring McInnes-Dobbie design. Concurrently, the pointer of the concealed-spring McInnes indicator was changed to the new Dobbie 'over-bar' layout.

A compact half size Dobbie-McInnes external-spring indicator, D1C-32988, dating from 1944 or 1945, used to indicate 'explosion' (internal-combustion) engines. Similar indicators were made in 'Large' and 'Small' versions for use with steam engines. Museum of Making collection.

The design of the pointer linkage identifies all Improved McInnes, McInnes-Dobbie and Dobbie McInnes indicators made since the beginning of the twentieth century. The improvement concerned the ease with which the spring could be changed. It was simply necessary to unscrew the top cap (carrying the pointer linkage) and lift it clear of the indicator body; the vulcanite locking collar could then be unscrewed and the spring removed. McInnes-Dobbie indicators were very successful, and are still commonly encountered. They had been renamed 'Dobbie McInnes' by 1909, though many instruments continued to be marked MCINNES-DOBBIE and misleadingly marked printing blocks (already some years old) continued to be used for some time.

Thompson, 1899

Among the first of the American exposed indicators seems to have been a modified Thompson type, patented in 1899 and introduced commercially by the American Steam Gage Mfg Co. by 1902-3. The spring was mounted on top of the platform, held in a threaded collar supported by a bar across two pillars. The two-leaf pointer lever was bent so that the piston rod could pass through a central aperture.

Similar indicators were also made by Schaeffer & Budenberg in Magdeburg (though often marked 'Manchester', where a limited company had been registered). The first type had the spring beneath the platform, connected to the piston by a bifurcated arm with prominent lightening holes. The piston housing was connected with the platform by three pierced steel straps. Instruments of this type were only made in small numbers, and had been replaced by 1908 by an improved design with the spring supported above the platform by a two-strap bracket. The side links were duplicated to allow the piston rod to pass between the stirrup-shape pencil arm. German-made Thompson derivatives were more successful than their American counterparts, which, with the exception of the Trill, lost so much ground to the Crosby type that manufacture had apparently ceased by 1917.


A German '1902 Model' external-spring indicator made by Dreyer, Rosenkranz & Droop of Hannover, probably in about 1907. Extremely well made and beautifully finished (though a cumbersome variant of the Thompson amplifying mechanism), this particular example has a wide range of accessories. Museum of Making collection.

Houghtaling-Tabor, 1900

An exposed-spring version of the Tabor indicator was introduced in 1900 on the basis of a patent granted to William Houghtaling, retaining the basic curved-track-and-roller amplifying system of the perfected internal-spring design. However, two cylindrical standards protruding vertically from the platform supported a bridge, and the spring was clamped between the bridge and the platform by tightening a screwed collar. The amplifying mechanism was offset in an auxiliary chamber alongside the piston cylinder, allowing the pointer to be activated by a 'lifter' on the piston rod. Much of the existing production tooling could be used and the springs could be changed exceptionally easily, but the unconstrained springs were prone to flex laterally; accuracy may not have been exceptional. In addition, the clumsiness of the design—which had much in common with the American-Thompson and the first Dreyer, Rosenkranz & Droop or Maihak patterns in Germany—suggests that the Ashcroft Manufacturing Company rushed the external-spring Tabor into production prematurely.

The external-spring Tabor indicator had a vertical plate alongside the spring standards, containing the cam-track regulating the pointer. Museum of Making collection.


The introduction of exposed-spring indicators by the Ashcroft Manufacturing Company (Houghtaling-Tabor type) and the American Steam Gauge Company ('Improved Thompson') persuaded Crosby that a competing design was needed. Neither of the rival designs were particularly efficient, and it must have seemed that a better instrument was easily developed. For this Crosby turned to a design by Theodore Davidson of Salem, Massachusetts. Assigned to the Crosby Steam Gage & Valve Company, U.S. Patent 713611—sought on 10th May 1901 and granted on 18th November 1902 — protected a variation of the standard Crosby indicator with the spring moved to encircle an extension of the piston rod carried above a bridge supported on two cylindrical standards. The central portion of the rod was split to allow the linkage to pass. Unscrewing and removing the threaded piston-rod cap allowed the spring to be changed in an instant.

The Davidson patent provided the basis for a number of Crosby indicators. The first, the 'Crosby New Engine Indicator', seems to have appeared about 1904. It followed the patent in virtually all respects, embodying traditional Crosby practice in the design of the amplifying mechanism and the 'spherical piston' associated with the earlier enclosed-spring designs. This was claimed never to jam or bind. The New Indicator was customarily supplied for use with steam engines, and had a piston diameter of 1.128 inches. The paper drum usually had a diameter of two inches, but a version could be supplied for use with high-speed steam and gas engines: this had a piston offering half the surface area of the steam-engine type—one half of a square inch instead of one square inch—and a stronger tracer mechanism. An all-steel indicator could be supplied for refrigeration machinery in which ammonia was used (the standard versions were nickel-plated brass).

By 1910, the standard New Indicator had been joined by the 'New No. 2 Indicator', a smaller and somewhat cheaper version of the original instrument intended for use with high-speed steam and gas engines. This was made with a piston diameter of 0.7979 inches (giving a piston face of one-half of a square inch) and had a slimmer body than the New Indicator; it also had a drum with a diameter of 1.5 inches, and could be supplied for use with gas engines with a piston-face area of only a quarter of a square inch (piston diameter: 0.564 inches). A minor variation allowed either 'steam' or 'gas' pistons to be used by providing exchangeable pistons and liners; it could be identified by a prominent hexagonal shoulder at the base of the body. Pistols with a face as small as one-twentieth of a square inch could be obtained to special order.

The 1910 handbook may have been the first to illustrate the 'Crosby [New] Indicator with Drum for Taking Continuous Diagrams'. Some years earlier, Crosby had experimented with two versions of the continuous recorder. Similar in principle but differing in detail, both patented in the U.S.A. in October 1907, they were the work of Theodore Davidson and Frank Wolfe. The Davidson version was preferred, as it allowed the space between successive diagrams to be altered by a ratchet-and-pawl mechanism.

The Crosby 'New Model' indicator was a distinctive design, made in accordance with a patent granted in 1902. This photograph shows a pre-1914 example in the Museum of Making collection, number 01393.

Production of enclosed-spring Crosby indicators seems to have ceased at the end of the First World War, though existing inventory allowed them to be sold 'as new' for several years thereafter. By 1927, however, only the exposed-spring 'New' and 'New No. 2' were being marketed (the former only in the guise of the continuous-diagram pattern). A rationalisation of Crosby's inventory system had gained them 'B'-prefix designations. The standard steam version was 'BC-101', the gas/oil-engine version was 'BC-201', and the all-steel ammonia instrument was 'BC-701'. The continuous-diagram indicator was known as 'BC-171'.

It is possible that the Great Depression that followed the Wall Street Crash of 1929 brought production of Crosby indicators to a standstill. They were still being sold in 1939, but it is suspected that no new production had been undertaken for several years. When the U.S.A. entered the Second World War at the end of 1941, it is perhaps no coincidence that the indicators supplied to the U.S. Navy and merchant marine were made by Trill (for use with steam engines) or by Bacharach (for the diesels).

The distribution of Crosby indicators throughout the work encouraged copies to be made. The most successful came from Germany. Only three manufacturers dominated the pre-1914 German market: Schaeffer & Budenburg of Buckau bei Magdeburg and Dreyer, Rosenkranz & Droop of Hannover made Thompson-type indicators with exposed springs, but Maihak made a variant of the external-spring Crosby with duplicated links and a ‘Y’-shape pointer that allowed the piston-rod extension to rise between its arms. This layout, which gave a symmetrical layout without the need to cut a slot in the piston-rod extension, had been patented by Wilhelm Lehmann in 1906.


The 1902-type Maihak combined the Crosby linkage with two springs placed outside the body, where they were extended by a transverse bar attached to the piston rod — an ineffectual solution to the problem of providing an exposed-spring instrument; and undoubtedly a production expedient. It scarcely lasted in production for more than a year, being replaced in 1903 by a much more compact design. The Crosby linkage was retained, but the tail of the piston rod was extended upward through a bridge (held above the body-cap by two rod-like standards) and then through the spring. The compression spring was held in place by a threaded nut.

This was in turn replaced by the Staus design (1905–8) and then the Lehmann-patent type made from 1907 onward. These all embodied a modified Crosby linkage with a bifurcated tracer arm and the links duplicated to allow the piston rod to rise through the centre of the compression spring. The principal difference was that the Staus-pattern Maihak retained the small incurved standard at the front of the amplifying mechanism, had a single back link, and carried the spring on a cap supported by two rod-like standards; the pressure applied by the trace to the drum was controlled by a threaded rod (with a small turned wooden handle projecting to the rear), which contacted the back of one of the standards. The Lehmann pattern had a short peg-like standard anchoring each of the duplicated front links, a bridle-type back link, and a compression spring bearing on an abutment forged integrally with the body plug. The trace-stop rod lay towards the front of the unit, bearing on a post on the centreline of the platform.

Left: Maihak Typ 50 indicator no. 20896 dates from the 1930s and shows evidence of hard use. The small cylindrical container protruding beneath the drum-nut is a Stauffer-type screw feed lubricator. Right: a Maihak-made Typ 50 indicator, dating from the 1950s, with its oak case and accessories. Author's collection.

The Staus indicator came in three sizes, large (grosses Modell), intermediate (Mittelmodell) and small (kleines Modell). Lehmann-type Maihaks were also offered in three sizes prior to 1914:‘Grösse 1’ with a 51mm diameter drum and a maximum speed rating of 300 rpm, ‘Grösse 2’ (38mm, 600 rpm) and ‘Grösse 3’ (30mm, 1500 rpm). Work on Lehmann-type indicators continued until 1985, when manufacture was then entrusted to Leutert GmbH. Though the current Leutert Typ 30 and Typ 50 indicators embody improved materials and simplifications in construction, they are essentially similar to the 'Original Maihak' instruments that were being made in the period between the world wars.


Though Dreyer, Rosenkranz & Droop made a few Crosby-like indicators in the 1930s, shortly before trading ceased, this was the only other German manufacturer to specialise in them. Founded in Altona in 1911, apparently to develop diesel engines, Lehmann & Michels soon began to make Maihak-like indicators in accordance with the 1906 patent granted to Wilhelm Lehmann. It is assumed that Lehmann had initially licensed his design to Maihak, but retained the right to exploit the design himself. Pre-1918 Lehmann & Michels indicators can usually be identified by a light bell-like housing protecting the amplifying linkage. Instruments of this type are still being made in Germany.

Left: 'Large Model' Lehmann & Michels indictaor, no. 5261. This particular example was sold in France soon after the end of the First World War and, to avoid anti-German resentment, shows no maker's marks. Right: a sectional drawing of a typical 1930s Lehmann & Michels indicator. Note the bell-housing protecting the amplifying mechanism, and the Stauffer-type lubricator on top of the drum spindle.


Copies of the standard Maihak Type 30 and Type 50 in-struments were certainly being made in the 1950s and 1960s in the German Democratic Republic (DDR) by Metallwerker AG of Meerane/Sachsen. In addition, Maihak/Lehmann & Michels-type indicators were made in Japan by Nagano KK until the late 1990s.

A typical British-made Casartelli indicator, probably dating from about 1910. This particular example has seven springs and two piston units (one for steam engines, one for 'internal-combustion engines'). Though the amplifying mechanism is similar to the McInnes and McInnes-Dobbie patterns, it is held to the cylinder cap by a large spiral spring at the rear. Author's collection.

James Atkinson (1846-1914) of 'Mellor, nr. Stockport [Lancashire], Engineer', was one of the best known of the British experimenters active prior to the First World War. His designs, which included the Differential Engine of 1885 and the Cycle Engine of 1886, were remarkable products of a fertile imagination. However, they failed to challenge the supremacy of the conventional horizontal-cylinder stationary engine. Atkinson was also responsible for modifying the design of the standard engine indicator to reduce the effects of the shock of combustion. Protected by a British Patent granted in 1906 (no. 5391/06),Atkinson's design shows Richards and Crosby-type indicators with a sturdy coil spring interposed in the links to 'register a more uniform and correct line as well as to relieve the gear from heavy strains or excessive vibration…' Indicators with Crosby-type amplifying gear were made by Casartelli of Manchester. They were similar to contemporaneous McInnes-Dobbie types, but could be easily identified by a large coil spring that rather clumsily held the rear of the amplifying mechanism to the platform. Indicators of this type were usually boxed with alternative steam/internal-combustion engine pistons to facilitate a dual-purpose role.

Simplex, 1894

A later embodiment of the pantograph will be found in the Simplex indicator, marketed commercially by Elliott Brotthers of London. Patented in Britain by Moses Wayne in September 1894, the Simplex is easily recognised by the 'sugar tong' or 'V' spring compressed between the top cap and the tip of the piston-rod extension. The spring could be removed simply by sliding it sideways. The piston unit could then be detached by unscrewing the locking ring and then lifting the piston and pointer assembly upward and out of the slotted tubular housing.

The patent showed a Darke-type elevator, but the pointer linkage of the production version was a classical four-lever pantograph which promised an accurate representation of pressures throughout its range without the need to ensure—as Ashcroft had done with the Tabor and Elliott with the Darke—that additional fixtures were required to direct the recording point in a straight line. The remainder of the Simplex was very similar to the standard Elliott made 'Improved Richards' indicator, which was still being made in quantity, and a modification of the Darke pattern detent slid in a small circular plate held to the piston housing with two screws.


The Elliott-Simplex indicator. The engraving (left) dating from 1897 shows what is assumed to be the prototype form, whereas the photograph of 'Pattern A' indicator no. 453.A (right) shows the perfected series-made type.

Vigorously promoted for the ease with which the spring could be changed and regulated, and the manufacturing advantages possessed by the distinctive spring ('lends itself to accuracy of calibration in any range', 'a simplicity of attachment [that] will recommend it to all engineers'), the Simplex was not particularly successful; production, apparently confined to1897–1901, was comparatively small. Two sizes were made: 'Pattern A' for normal speed ranges, and 'Pattern B', with a small drum, for high-speed use. Serial numbers will be found suffixed with the appropriate letter, and the springs were similarly distinguished to ensure they were used with the correct indicator-a typical example reads 'A' over '1/150' and '+375', showing a spring requiring a pressure of 150lb/sq.in to compress it by an inch and that 375lb/sq.in above atmospheric pressure was the maximum permissible in an 'A' class SImplex.

It is suspected that Simplex instruments were numbered in the same basic series and that no distinction was drawn between the two sizes other than the 'A' or 'B' identifier. If this proves to be true, then less than a thousand units were ever made. Sales were comparatively slow, as most British purchasers – still very conservative – preferred the Elliott-made Richards indicator. The highest-number Simplex yet found (no. 913.B) bears the marks of 'Elliott Bros. (London) Ltd', a change in corporate structure than does not seem to have occurred until shortly after the end of the First World War.

The problems with the Simplex are said to have concerned the excessive inertia of the heavyweight pantograph assembly, though the unique tong-springs may not have been popular. They may not have been as durable as the standard coil type, and regulating them may have been much more difficult than Elliott Brothers claimed. in addition, though the lighter springs were easily detached, changing the stiffer patterns required considerable manual strength or the assistance of a compressor.

A comparison between the original 1897-vintage engravings and the surviving instrument shown in the photograph shows that the perfected version had a simplified lightweight folded-metal pantograph that was pinned together instead of heavy links retained with screws, and the design of the piston-support bracket was altered to improve rigidity. There are many other changes in detail, which suggests that tests undertaken either with a single prototype or a 'pre-production' batch revealed shortcomings that Elliott took pains to correct.


External-spring indicators were more suited to internal-combustion engines than the enclosed-spring types, owing to the extreme temperatures encountered in the cylinders. An attempt to provide a comparatively conventional indicator suited to high-speed car and aeroplane engines was made in the 1920s by Englishman Charles Gale, who received British Patents 192551 and 192578, each sought in 1921, and US Patent 1483171 granted in 1924. Externally, Gale's indicator looked to be a conventional design with the spring mounted externally on a large bracket and an oscillating drum. The principal novelty lay in a valve controlled with a small hand-wheel, which allowed pressures in the engine to be monitored at differing positions in the operating cycle. A dash-pot or comparable damping system eliminated unwanted vibrations.

One method of overcoming the vibration problems encountered in high-speed running was provided by the micro-indicator, accepting that small-but-perfect diagrams were preferable to large traces compromised by undulations. The earliest attempt to provide an instrument of this type was made in Germany by Otto Mader, whose 1912 design was made in small quantities by Gebrüder Stärzle of München (Munich).

Mader's indicator consisted of a platform-like body containing a steam-cock and a small piston that could slide vertically against the resistance of a co-axial coil spring. A small 'writing lever' slid laterally on top of the piston rod as the short rocking-bracket to which it was attached was moved by a suitable part of the engine. A constraining spring anchored to the platform prevented the writing lever disengaging unexpectedly while a small stylus drew up to 24 minuscule diagrams (only about 2mm high) on a smoked-glass plate that slid in channelled supports in the rear of the indicator body. The 'cards' could be fixed by immersing the glass in Canada balsam, and viewed either with the aid of a low-power microscope or by enlarging them photographically.

Left: the Mader micro-indicator of 1912. Right: the Collins micro-indicator of 1922, showing the diagram disc (greatly enlarged). Drawings from K.J. DeJuhasz, The Engine Indicator (1934).

The 'Collins Micro-Indicator for High Speed Engines', patented in Britain in 1922 by W.G. Collins, director of the Cambridge & Paul Instrument Co. Ltd of London, took another approach. A very compact design, relying on a very small piston and a sturdy riband spring to move the stylus that engraved tiny diagrams on a small ratchet disc, the Collins design was praised by the British periodical Engineering: 'The moving parts are light and they move through small distances; there is little inertia, an entire absence of linkages and uncompensated joints, and there is, therefore, small liability to wear. With a piston area of 1/4 sq. in. the natural period of the recording system is about 1/1100 of a second.'

The diagrams produced by the Collins instrument were only about 3mm long, with a maximum height of about 2.5mm. Though they were difficult to interpret with the naked eye, a simple tubular microscope with a graticle allowed the operator to read the dimensions of the diagram to the nearest hundredth of an inch. The raised edges of the trace helped to make the diagrams surprisingly distinct.

Ten diagrams could be taken in rapid succession from high-speed engines, but the Collins design never prospered. This was partly due to the ready availability of conventional indicators, which could handle most of the day-to-day work of a travelling engineer; the need to have an electrical supply at hand, even if only a battery, and the impossibility of using a planimeter to interpret the micro-diagrams were considerable drawbacks. Though a few Collins micro-indicators were sold commercially in the 1920s, they soon faded into obscurity in the face of competition from multi-cylinder analysers such as the Farnboro pattern marketed aggressively by Dobbie McInnes.

Another type, the work of the German engineer Wolfgang Pabst, was used by the Deutsche Versuchsanstalt fur Luftfahrt from 1929 onward. Pressure on a diaphragm raised a lightweight rod attached to a diamond-point stylus, which engraved a diagram on a ground-glass slide that was oscillated during the engine cycles and moved sequentially by clockwork. The diamond allowed a very precise trace which could be as fine as 0.002mm: ideally matched to the diagrams, which were usually only about 0.5mm high!