Perris lathes were in production, in various guises, for just a dozen years until the untimely death of engineer Brian Perris in 1976.  It is well-known that his final design, the cantilever bed SL90 lathe, remains in production today as the beautiful little Cowells lathe.  What is not so well known is that the brief history of the preceding Perris PL Series of lathes was complex and fascinating.  Here we consider this chronology and, subsequently under separate headings, we shall examine details of the various machines.
Comprised, principally, of five variants on a common theme, the PL90 Series was a tiny machine with just 6 in. between centres, plain-turning and fitted with a simple cantilever bed. The PL90 lathe was joined in 1966 by the PL100, a sophisticated little BGSC lathe with 7 in. between centres and a box bed.  The PL100 also received a very good review, in the June 1966 Model Engineer.
The first Perris lathe to be introduced was the PL90 type 'Ace Minor', supposedly a product of a marketing company, Urquhart Machine Tools Ltd. of Clerkenwell, under whose label it received a glowing review in the January 1965 Model Engineer. However, the same machine was sold as the Centrix 'Micro', ostensibly the product of the 'Centrix Works' of Centrix Precision Products Ltd., Shoreham-by-the-Sea, Sussex (which co-incidentally, was where Brian Perris started his own engineering firm in the early 1960)s.  Despite the distributors' assertions that rebranded Perris lathes were their own products, Perris Engineering built them all, initially at Shoreham and then at 4 and 5 Meadow Rd. Industrial Estate, Worthing, Sussex. As we shall later see, Perris Engineering relocated twice again over the following decade.
With a 0.5" x 16 t.p.i. BSF mandrel nose and two 0-MT taper sockets, the three PL90 variants were badged as the 'Ace Minor' and 'Centrix Micro': (a) plain-turning, (b) back-geared but without a powered leadscrew, and (c) 'horological.'  The latter was wishful thinking for its distinction was a collet-holding spindle with unfortunate and significant design limitations.  The two PL90 variants sold under the Perris name were called the 'PL90 Standard' and the 'PL90 Special', collectively known as the 'Perris Model Maker's Lathe', and characterised by a finer (0.5" x 20 t.p.i. UNF) spindle thread - the re-badged versions having the same two No. 0 Morse taper sockets but a 0.5" x 16 t.p.i. BSF nose. 
From the outset Brian Perris reserved unto himself a superior 'Special' type of PL90, elements of which were carried over into his later SL90 lathe (and persist in the now familiar Cowells lathes produced today).  The 'Standard' version of the Perris PL90 'Model Maker', priced for the modeller with shallow pockets, was simply the regular type Ace and Centrix lathe. 
In the PL series, screwcutting was unique to the PL100 model, which was advertised as the Perris 3-1/2" Screwcutting Lathe.  The '100' seems rather fanciful since, apart from its between-centres distance its principal dimensions were the same as the PL90, the "100" coming from the diametral swing over the bed.  The original PL90 lathe, with just 6" between centres, did not include a screwcutting variant. 
Despite their budget pricing, the re-badged PL90s, the Ace Minor and the Centrix Micro, were as rigid and accurate as the Perris brand lathes.  The difference was features rather than quality.  Apparently the re-badged PL90s were only sold in assembled form during the few years when they were available.  During that time the PL90s sold by Perris Engineering were only sold as kits, although from the late 1960s they could also be bought assembled.  One suspects that previous distribution arrangements had disallowed Perris Engineering from selling assembled lathes other than the PL100, a machine unique to the Perris name.  The prices between the rebranded PL90s and the Perris PL90 kits were comparable, so with the Perris the purchaser got superior features but did not pay for the assembly or painting.
Exemplary assembly instructions, plus well-finished and consistent components, provided the novice with a morale-building and technically valuable learning exercise.  Self-assembly also saved a third % on the purchase price, but for this the owner also had to clean the castings with spirits and apply the paint provided. The instructions for the later SL90 machine included advice on perfecting the alignment for precision work.  If the purchaser lacked the necessary Verdict Junior dial indicator, said the instructions, the accuracy should nevertheless meet the average modeller's requirements.  Indeed it did, for all of these machines were carefully made with ground slides, bed, and spindle.
The PL90 Series lathes were well-regarded and widely sold.  The Ace and Centrix rebrands were carried by various tool merchants across the UK.  Known distributors of the Perris brand lathes included Bond's O' Euston Road and  Buck & Ryan in London, as well as numerous forgotten, local-engineering supply firms such as P.W. Stephenson's in Darlington.  It is nowadays hard to imagine the proliferation of small tool distributors that has been swept away by superstores and mail-order.  The Perris factory exported them directly to overseas purchasers, and there were at least a few overseas agents, including Ford in Toronto and Pascal Hardware in Ottawa.  The Model Maker's 'Special' illustrated on this page was obtained, assembled, in 1971 from Pascale. 
The PL series was succeeded, in the early 1970s, by the Perris-designed, cantilever bed BGSC lathe which continues to be made, in improved form, by Cowells.  Initially Brian Perris promoted this new machine as the 'revised' and 'improved' PL90 Model Makers Lathe.  This nomenclature would have been fine had not the 'old favourite', the original PL90 Model Maker's Lathe, remained in production for another year or two.  To further confuse matters, early (1973) promotional literature for the new lathe carried illustrations of the old lathe with a description of the new: The Modelmaker - All metric or all inch.  An old favourite the Perris PL90, back in revised form.  A bit more capacity with optional backgearing and autotraverse. 
Eventually Brian Perris remedied this confusion by renaming the new lathe the 'SL90' and completing the retirement of the previous machines.  We are using the latter-day Perris terminology 'PL' and 'SL' to distinguish between the old and new series of Perris lathes in order to avoid perpetuating this perplexity.
Retiring the PL Series was clearly the designer's intent.  The SL90 was a small machine, just three inches longer than the diminutive PL90 'Special', but able to punch above its weight in terms of capacity.  In particular, its centres distance was 8 in., compared with the PL90's 6 in. and the PL100's 7 in., and it had an excellent long-travel, three-slot cross-slide developed from the PL100.  There was little point continuing production of the PL series because the SL90's versatile bed casting offered a common platform for plain-turning, back-geared, and BGSC variants according to the buyer's depth of pocket.  There also is good reason to think that Brian Perris knew that he had designed an enduring class of machine with great potential for modification.  Certainly it allowed for streamlining the production line and rationalising a spares situation which a diversity of versions, marks, upgrades and special-order features had made complicated.   Yet perhaps the main reason for adopting a common platform was that the new lathe's general superiority in specification rendered the PL90 and PL100 machines much less marketable.
The transformation of the Perris product line was challenging for reasons beside the promotional faux pas referred to above.  Indeed the early 1970s were a difficult time for Brian Perris.  In the background was his involvement in the design, marketing, and possibly manufacture of other machines suspected to include the Sartglen Pixie, the Hector, and the recent incarnation of the Flexispeed.  We do not know how profitable these side activities were, but we do know that his entrepreneurism and workaholic nature contributed to great personal and financial stresses.  Supply problems complicated the launching of the SL90.  Besides having to recycle previous promotional literature, re-stamped in blue with a new works address, not all of the components were ready on time.  To illustrate this, the SL90 was initially (1973 and 1974) advertised as 'auto-traversing' only, and where screwcutting facilities are needed, all that will be required will be a few extra gear wheels and these will be available at a later date.
There was apparently some disruption of the supply of his lathes while the works re-tooled for the SL90 and relocated from 61 George St. to 10 Sprowston Road in Norwich.  This seems to have reduced his existing distributor base, making direct sales more important, and requiring advertising funds which he did not have at a critical time. 
When Perris Engineering resumed advertising, at a minimal level, the SL90 was already known in model engineering circles from his attendance and the Model Engineer Exhibition but its presence in showrooms was quite limited.  It was unfortunate that, despite vigorous exhibition attendance by Brian Perris, the Model Engineer did not carry a review of the new lathe.  Possibly this was because, while he had many friends in the model engineering world, he did not have the cash to put a half-page advertisement in the Model Engineer concurrent with a review article.  By the early 1970s the advertising space in this important journal had diminished, adverts were premium priced, and the overall hobby press was more expectant than ever that reviews would be accompanied by advertising revenue. 
Perris Engineering ended abruptly when Brian Perris died - presumably in motor accident - on his way home from the 1976 Model Engineer Exhibition.  It has been remarked that this was not surprising considering his pace of life and degree of stress, but in any case, the UK was deprived of its best-known and probably most successful designer of small engineering tools.  By then, the SL90 was the only Perris lathe in production although spares for the PL series were obtainable, and a few complete PL90s and PL100s still persisted in the showrooms of tool distributors.  The firm went into liquidation immediately.  Sid Cowell of Norwich swiftly bought the assets and patterns, putting the SL90 back into production in time for the 1977 Model Engineer Exhibition.  The first Cowells 90 series machines appear to have been identical to the Perris SL90 but for 'Cowells' cast into the bed, and blue enamel replacing the green crackle finish which had distinguished Perris lathes.  Support for the PL90 and PL100 lathes was discontinued although some current Cowells accessories will fit a PL90, PL100, or SL90. 
A more complete story of the Cowells can be found in the Cowells entry in this Archive.

The PL90 Lathe
Appearing first in 1963 the lathe was not badged "Perris", but as the 'Ace Minor Model Maker's Lathe' of Urquhart Machine Tools Ltd., Clerkenwell.  The January 1965 Model Engineer had a favourable review, by Edgar Westbury, which is worth quoting at length:
Superficially, the lathe bears a general resemblance to others which have been produced in the past.  It has several improved features, and a higher accuracy and finish than most lathes of comparable size which we have seen in recent years.  The design follows engine lathe practice, as distinct from the practice more common for clock and instrument lathes.  It incorporates a cast iron box bed with a flat top, gap, and dovetail sides.  A separate headstock is fitted, with parallel half-split bearings for the mandrel, which is hollow and takes No. 0 Morse centres.  The nose is screwed 1/2 x 16 t.p.i. [1/2 BSF] and a three-step V pulley is provided.
A fully compound slide is fitted to the sliding saddle, which is traversed by a lead screw with a two-start thread of 10 t.p.i. lead.  The tailstock has a set-over adjustment, a lever clamp, and a hollow barrel with No. 0 Morse socket.  A maximum length of 6 in. (150 mm.) is admitted between centres, and the centre height is 1 3/4 in. (45 mm.).  The swing in the gap is 4 1/2" dia. (120 mm.) and the cross-slide is 1 7/8 in. dia. (48 mm.).
Standard equipment consists of a 3 1/2 in. faceplate, a 2 in. catchplate, two point centres, a swinging countershaft with three-step and primary drive pulleys, and a V belt.
Starting at almost the same time, this Perris lathe was also sold as the 'Centrix Micro', supposedly the product of Centrix Precision Products Ltd.  The name 'Centrix' is remembered mainly in connection with a mid-sized, cast aluminium, multi-purpose woodworking lathe reviewed by the Model Engineer in 1947, a useful machines for the amateur woodworker.  The surviving Centrix woodworking lathes appear geriatric in both architecture and age, which with an absence of further mention of the lathe or the builder, suggest that Centrix wood lathes were out of production by the early 1970s if not sooner. 
Perhaps there were other re-badged PL90s besides the Ace Minor and Centrix Micro, but re-branded PL90s were not on the market long.  The last Micro is thought to have left the so-called Centrix Works in February 1966, co-incident with the Model Engineer's favourable review of the PL100 and relocation of the Perris works from Shoreham to Worthing (later to Norwich).  Regardless of the selling label, it did not take long for the model engineering world to realise that Brian Perris was the designer and producer of the entire PL range of lathes.  The Perris lathe, as the type was soon known, was markedly superior to the familiar Flexispeeds, Lanes, Grindturns, and Adepts which the past generation of modeller's had known.  The boastful Ace Minor advert, shown here, was factually correct.
Apart from coarser BSF mandrel nose thread, the basic model 'Ace Minor' and 'Centrix Micro' were architecturally similar to the 'Standard' Perris PL90 lathe.  The former were distinguished by colourful, lacquer-based adhesive transfers whereas the Perris brand machines had natty engraved plates which, being glued on, tended to come off.  The photos in the Centrix section clearly illustrate this basic model of PL90 lathe.
The differences were much more apparent between the basic and the upgraded models.  To continue the Model Engineer's review of the Ace Minor: Special versions of the lathe for clock-making and second-operation work are made, with appropriate accessories such as collet chucks and, together with other refinements.  The clockmaking version was rather less horological than one might expect.  It was simply an economy version of the collet-holding Perris PL90 'Special', a machine we shall return to presently.  Very few of these re-badged collet-holding lathes are believed to have been sold, whereas the Perris brand 'Special' was popular and a number survive in good condition today. 
The 'second-operation' specie of PL90 was a backgeared version of the standard product.  This was directed at model engineers needing to turn large iron castings, such as locomotive wheels, but lacking the money for a full screwcutting lathe.  Shown here is an advert, from the 4 February 1966 Model Engineer, depicting the Centrix version of this lathe.  The back-gear option required a slightly longer bed in order to accommodate a lengthened headstock casting with the back-gear assembly.  A pressed metal guard, shown here removed, enclosed the spindle unit in a manner reminiscent of the later SIMAT.  The back-geared PL90 could only be had assembled and to special order.  Although Perris Engineering continued to offer this version after the Ace and Centrix brands disappeared, it could not have been popular since there was scant mention in the modelling press and examples have not come to light.
Directed at the likes of railway modellers, who wanted a plain-turning, traditional centre lathe, the 'Standard' type of PL90 was an improvement in quality compated with what had been available as the Adept, Lane Wizard, Flexispeed and so on.  The 'Special', offered only as a Perris, was conceived as a deluxe version of an already superior machine with the most obvious difference being a superior top-slide mounted on a tapered spigot similar to a Myford ML-7 or a South Bend - and very like that on the current Cowells.  Conversely, the 'Standard' lathe employed the primitive lug-and-bolt arrangement favoured for a half-century by makers of cheap British small lathes.  This system required frightful torque lest the top-slide rotate under load with concomitant negative impacts upon the workpiece.  This nuisance was partially mitigated by having a lug with a slot, not just a hole, by which two bolts could secure the casting to the longitudinal T-slot.  This would have been effective for parallel turning but only the most gradual taper could be attempted without removing one of the bolts and risking angular displacement of the slide under load. However, the 'Standard' did have a useful cross-slide boring table with no fewer than three T-slots, a quantum leap compared with what else was on offer amongst modeller's lathes.  In utilising a spigot-mounted top-slide, the 'Special' PL90 had to dispense with one of these slots, making its table less useful for boring or vertical milling.  Its main purpose was to mount the effective little travelling steady and back tool post offered by the manufacturer.  It must be remembered that the high-end PL90 did not masquerade as a full-function lathe.  For that purpose the PL100 was available.
  Other highly desirable refinements were included in the 'Special' PL90 with the 'Standard' model's  un-graduated aluminium wheels replaced with steel versions crisply engraved in 'thous'.  The cross-slide wheel had a natty zero-adjust micrometer sleeve and also the superior tailstock fitted to the PL100 (and perpetuated today on the Cowells) - this dispensing with the through-bored handwheel type, fitted to the 'Standard', in favour of a dead-length, self-ejecting type which again was reminiscent of the Myford ML-7.  More importantly, Allen head grub screws were provided for set-over, replacing a less satisfactory simple tenon-and-vertical-bolt fixing arrangement. 
The spindle was the final differentiating feature of the high-end PL90: like the economy model it was made of good chrome steel and was carefully ground.  It also ran in plain iron bearings which were given a fine internal finish.  The 'Special' PL90's spindle was dimensionally different.  Firstly it had a refreshingly long extension at the rear, suitable for use with a drawbar when collets were used, and long enough for user modifications like a home-made index plate.  Secondly, the 'Special' had a 0.5" x 20 t.p.i. UNF mandrel nose, bored 8 mm for collets and with a truncated internal cone.  This was the same mandrel nose fitted as standard to the PL100 lathe, and was perhaps the best-known distinguishing feature of Perris lathes.
Mandrel noses of this type were not new.  They appeared first on the CAV Wade lathe of 1924 and reappeared, in 1954, on the Unimat DB/SL.  A similar arrangement continues with the Taig/Peatol lathe.  The main advantage for most manufacturers has been cheapness of manufacture, but this was not the starting point for Brian Perris, who had collets much in mind.  He employed a similar mandrel nose Pixi/Jason lathe for which were supplied dead-length compression collets closed by a threaded collar which acted upon the nose thread.  For the PL90 and PL100 he supplied conventional draw-in collets closed by a tubular drawbar.
Unfortunately the small 0.5" nose thread disallowed an internal cone large enough in diameter to fully accommodate the cone of a standard 8 mm horological collet.  A regular collet fits the Perris bore perfectly but it protrudes considerably, and tightening the drawbar would impose uneven strains sure to damage a regular horological collet.  For this reason Perris Engineering supplied special draw-in collets with a reduced cone diameter.  These were available in sizes 3/32", 1/8", 3/16", and ¼", plus a slotted but blank which the user could bore to any size needed.  It is hard to call such a limited selection 'horological' but these sizes did hold the diameters of commercial rod and cutters likely to be needed.  It has been reported that dead-length compression collets, as per the Pixi lathe, were also available for the PL90 and PL100 but the existence of these has not been confirmed.  Note that some PL90s, and most PL100s, had a locating pin to engage in the slot of an 8 mm collet.  In practice, considering the light work undertaken, the absence of such a feature tends to go unnoticed.
Obviously, Perris in time realised that the tiny nose thread was a bad idea for an 8 mm collet-holding spindle, since the later adoption of a larger (14 mm) thread on the SL90 lathe permitted true WW-type horological compatibility.  A little 0.5" nose did, however, nicely accommodate the 11 mm diameter cone of a 6 or 6.5 mm horological collet,  and a few Perris lathes have spindle to this specification.  These are either user modifications or some of countless custom alterations that Perris Engineering was willing to undertake at a reasonable price.  Perris was doubtless aware that the smaller type of collet was rapidly falling out of use, although to the enthusiastic lathe owner, used specimens were available cheaply as a consequence of the rapid shift from mechanical to electronic watches.
Many owners of the PL90 'Special' did not bother with collets although Perris-made collets are commonly encountered with PL100s.  Often they made do with the manufacturer's screw-on drill chuck for which a threaded 0-MT adaptor was supplied for use with the tailstock.  Illustrated here is an original Perris live centre protruding from a PL90 mandrel nose.  This simple, renewable article is made of soft steel and the user was advised to make a skimming cut before each use.  As per the Unimat, the short cone would provide sufficient purchase prevent rotation during a truing operation, but a greater application of force would have it spin in the bore with damaging results. 
Lacking a taper socket in the spindle the home mechanic was severely challenged to manufacture taper tooling for the tailstock, but the PL90 buyer wanting a taper mandrel could simply get the 'Standard' model, with the UNF nose thread but with a taper socket, or request the same spindle on a 'Special'.  Furthermore, cylindrical arbors and holders with a simple cone are easy to make on a plain lathe, and it is straightforward to tap the end for a drawbar. 
A mandrel nose of 0.5" BSW or 0.5" UNF limited the depth of a Morse taper socket.  Like most lathe makers who used a half-inch spindle, Brian Perris had to employ a shortened 0-MT taper based on the smaller two-thirds of the Morse plug gauge.  Full-sized 0-MT tooling will fit into a PL90 but it protrudes considerably.  The current production Cowells taper appears identical to the truncated Morse introduced by Mr. Perris.  No other current production truncated tooling appears compatible, including the Sherline taper, which is based on the big end of the taper gauge.
Well-provided with basic accessories, those for the PL90 were mostly were compatible with the PL100.  Those accessories available in 1964 were a 3-3/8" diameter faceplate, four-jaw and three-jaw chucks fitted to be mounted directly on the mandrel nose, centres (dead, live, female and half), an angle plate, a vertical slide, a three-point steady, a tailstock die holder, and a drill chuck.  This was expanded to include a travelling steady, back tool post, Slocomb centre-drill holder for the tailstock, machine vice, indexing head, circular saw attachment, and four-tool turret tool post.  Most 'Standard' and 'Special' PL90 were fitted with a clamp-type English tool post.  Towards the end of production the 'Special' came with a single-tool block tool post similar to the basic Cowells item today. 

The PL100 Lathe
It is first necessary to understand that the PL100 appears larger in photos than it actually was.  Indeed, it echoed the full-featured Myford ML10 but had the centre height (1-3/4 in.) and shears of the little PL90, with just 7 in. between centres. 
Making its debut approximately 18 months after the PL90 had appeared on the market, the PL100 receiving a particularly positive review in the 1 July 1966 Model Engineer.  It is worth quoting this at length:

A new model engineer's lathe is always a highly welcome addition to the range of light machinery available to the amateur.  Our two pictures show the latest "Perris" lathe, a 1 3/4 in. centre machine designed especially for the model engineer who requires a machine with all the refinements usually associated with larger machines, such as back gearing, ground bed, steel gears with bronze bearings, compound slides, four-tool turret, back tool post, etc.
The Perris lathe swings 3 ½ in. over the bed, 2 in. over the cross-slide, and 4 ¾ in. in the gap.  Distance between centres is 7 in.  Eight speeds are obtainable using the standard pulleys; the ungeared range is from 282 to 1,000 r.p.m. and the geared range from 60 to 200 r.p.m.  A full range of threads may be cut with the change wheels supplied and a fine feed of 0.003 in. per revolution.
The headstock bearings are of phosphor-bronze and provision is made for adjustment.  The spindle is precision ground and is bored 5/16 in. through.  Collets with 8 mm. Body will fit directly into the nose without an adaptor.  An auto-cut-off to traverse is fitted.  The leadscrew is 16 t.p.i. ACME form and the handwheel is graduated in 125 divisions.  The saddle is fitted with proper gib with adjustment for wear.
Standard equipment includes a 2 3/8 in. independent chuck with reversible jaws, a 3 ½ in. dia. faceplate, a set of change wheels, a four-way tool turret, a headstock vee belt and matching pulley, a set of six Allen keys, two spanners and two centres.  A universal dividing head is available as an extra.
This small lathe is well designed, well made and nicely finished, and should prove most useful for model engineers and model railwaymen, and for light turning and instrument work.  It is obtainable from most tool dealers.

It is unclear from the above that the regular PL100 mandrel nose had the same 1/2" x 20 t.p.i. UNF mandrel nose as the PL90 'Special' and was also bored 8 mm for the Perris style of collet.  The PL100 was also available with a UNF nose carrying a Morse taper socket. 
The carriage assembly was generally similar to that of the PL90 'Special' except in two respects.  Firstly, the PL100's top slide was more refined and had slightly longer travel.  Secondly, the PL100 had a two-slot cross-slide table, resembling that of the PL90 Special, but with an extended feedscrew bracket similar to the Myford 7 series.  The extended bracket afforded sufficient inward travel to use the vertical slide, whose spigot fitted into the hole vacated by the compound top-slide.  The PL100's vertical slide did fit into the PL90's 'Special's table, but this arrangement was nigh useless because the PL90's cross-slide would not move sufficiently inwards.  This explains why the PL90 had a more conventional, non-rotating vertical slide secured to the table with T-bolts. 
Some have stated that an extended three-slot table was available for the PL100 as an optional upgrade, like the similar item provided for the Myford ML-7, but this has not been confirmed.  Nevertheless Brian Perris recognised the desirability of a long-travel three-slot table, more capable of traditional saddle boring operations, when he designed the SL90.  In order to achieve this he replaced the extended feedscrew bracket with a lengthened version of the table originally fitted to the PL90 'Special', thereby omitting the PL100's extended bracket.  This table remains standard on the Cowells 90 series.
A final distinguishing feature of the PL100 was etched maker's plates that were riveted, not glued, onto the machine.  These included a screwcutting chart inside the hinged backgear cover.

The SL90 Lathe
The familiar Cowells 90 lathe is essentially the SL90, the second generation of Perris lathe, and a brilliant fusion of the PL90 and PL100.  Dimensionally it is a PL90 with 8 rather than 6 inches between centres, a wider cross-slide and table and with the back-gearing and screwcutting equipment inspired by the PL100.  Since most of the history and development of this lathe occurred subsequently during Cowells ownership, and we have already touched upon origins of the SL90, our discussion here will be brief.
When Brian Perris died in 1975 the production of PL Series was over.  The SL90 was the only product, available in kit or assembled form, with four options: (a) plain-turning, (b) back-geared, (c) back-geared and auto-traversing, and (d) full BGSC.  After considerable difficulty the basic SL90 was in full production with plain bearings, a 14 mm x 1 mm mandrel nose, and a 0-MT socket.  A planned horological version, with conical bearings and a proper 8 mm spindle, does not seem to have made it past the prototype stage.  The writer would be grateful for any information on specialised versions of the SL90 which pre-dated the acquisition by Cowells.
Brian Perris also intended to produce a self-contained SL90 with integral electric drive.  Again it is not sure whether this proceeded beyond the prototype stage before his death.   Self-contained lathes were definitely a priority for the new owner.  The first self-contained Cowells lathes, on a sturdy pressed metal stand and a belt cover, appeared about 1977.  Cowells soon set about making a range of sensible improvements which likely would have occurred to Mr. Perris had be still been around.  Early changes were a clutch lever to replace the original knurled wheel and improved handwheels with large balanced cranks.  Another early change was to give the bed and tailstock slightly more angular lines.
During most of their production period the Perris lathes were oriented towards the model maker and model engineer.  Value for money was strongly emphasised.  The SL90's business model was modular, allowing the buyer to purchase just what was needed, with the option of upgrading later.  It is only at the very end of production that Mr. Perris put more emphasis on the high-end, premium nature of his products.  Cowells carried this further, swiftly dropping the kit option and stressing the precision nature of the machine, producing higher-standard versions which were definitely in the clockmaking class.  The firm's advertising increasingly gravitated towards horologists and others needing precision.  Tolerances were tightened and new accessories, such as a Jacot attachment, were introduced to this end.  Cowells eventually switched all the threads to metric but otherwise the machine is dimensionally similar to the SL90.  It reflects well on Brian Perris that his SL90 was designed with such upgrade potential in mind.
In the final analysis, we must remember that the intelligent acquisition of Perris Engineering allowed Cowells to carry on as the last producer of miniature lathes and related equipment in the UK.  It is prudent to bear in mind that the Cowells is the sole survivor, that it is expensive to manufacture in the UK, and that there is no remotely comparable lathe in production elsewhere today.  Such small and specialised firms will not survive without the patronage of quality-oriented miniature engineers.  These considerations might be kept in mind when contemplating the purchase of Cowells accessories for a Perris lathe.

Metric Perris Lathes
The PL series machines were designed to imperial specification including dimensions, mandrel nose thread, motion screws, and the fasteners which were BA.  The only known metric dimension in regular PL90s and PL100s was the 8 mm mandrel bore which corresponded to the body diameter of standard horological collets.
Metric versions of some PL series lathes were available from an early juncture.  After about 1967 the PL90 Model Maker's 'Special' and the PL100 could be had with metric top-slide and cross-slide feed screws and indices.  Like the metric Myford Series 7 lathes, the lead screws remained imperial but the leadscrew handwheels were graduated in approximate metric increments.  This was sufficient for most applications and the PL100's translation wheel permitted the cutting of metric threads.  A very few of the last production metric model PL100s had a proper metric leadscrew, requiring no translation gear to cut metric threads.
Also designed as an imperial machine, the SL90 was available with imperial or metric feeds and dials.  However, from the outset, it sported as standard the 14 mm x 1 mm nose of the Unimat 2, with a 0-MT socket or bored for 8 mm collets.  This larger thread diameter allowed a larger internal cone so that, finally, regular WW collets could be accommodated to the full extent of their cone diameter.  Apparently the 14 mm thread was the only option for the PL90, with Morse socket or fitted for collets.  The collet-only mandrel nose is thought to be a Cowells innovation but this is not certain.  It does seem certain that the SL90 type developed into a fully metric machine during Cowells manufacture, although the principal dimensions such as centre height are metric expressions of the original imperial units.

Copyright: Andrew Webster--Canada