Screwcutting on Pittler Lathes depends upon matched sets of worms and wheels which, for many years, have been in short supply. These are to be one of the first items the group hopes to produce - however, if you are in a rush to make your own you might like to know that the B2 used a set of 17 worms and wheels (able to cut 27 pitches) of 20 diametral pitch, with a pressure angle of 14.5 degrees.
The single-start worm was of around 1.030" O/D with a pitch of 0.1571", the five, ten, and fifteen-start worms were of the same helix angle (different to the single start) but further data is only relevant to the ten-start worm, of 2.880" O/D with a pitch of 0.160".
For the B2 Pittler the standard set of worms and wheels comprised:
(Multi start) 20,22,24,26,28,30,32,34,36,38,40,48,60
(Single start) 50,80
The full set comprised of 17 wheels, data has left the matter unresolved at present, although it suggests that certain numbers of teeth were duplicated in different starts, for example, I suspect that a single-start 30T gear was produced but this needs to be clarified.
There were three worms supplied with the standard kit, a fourth has been found, by one Pittler enthusiast, but is of little real use. These were-
Single start (standard) relating to fine feeds and pitches to 400 T.P.I.
Multi start (standard) relating to screwcutting pitches
Ten start (standard) relating to screwcutting pitches
Fifteen start (non-standard)
The machine relied on the bad practice of using different-start worms with the same worm wheels, a system possible only because great care was taken to maintain an identical helix angle between all of the worms; the wheels themselves were cut with a throat radius to correspond with the largest o/d worm to be used. The diametral pitch was 20 d.p. (identical to that used on the eries 7 Myford lathes) and the pressure angle would originally have been 14.5 degrees - though this does not matter so long as the same p.a. is used throughout the set. The worms were made from steel, the wheels from cast iron. The linear pitch of the one worm has been measured at 0.156" - which, if correct. would make for relatively easy reproduction,
Method of Production:
Unfortunately, the worms and wheels cannot be made from scratch on the machine itself, unless a full set is present. There was a wormwheel-hobbing fixture available but, unfortunately, they are very rare.
The course of action is as follows - a fixture must be produced, an example of which I am presently designing, that comprises a rotating fixture mounted on the lathe cross-slide of a conventional lathe (the Drummond in my case) based around a worm and wheel of 32 teeth and two starts. Using the standard Drummond changewheels, plus one 32-tooth wheel, it is possible to gear the fixture to the headstock mandrel at an appropriate ratio (angular movement of the fixture and the blank to be cut per revolution of the cutter bit driven between centres) for each and all of the wormwheels required. As the fixture and the cutter will be rotating as the teeth are cut, a straight-sided form tool, set to match the helix angle, and shaped up like an acme tool (for 14.5 p.a.) will automatically generate the involute profile that is required. The gearing must be geared up again, ten times for the multi-start wheels, and for this reason it is advisable to drive the lathe from the quadrant as opposed to the lathe headstock. The ratios are as follows - (trust me on this one, I wasted an entire August Bank Holiday doing the calculations) ...........

No. Teeth .… Ratio .… Simple gear train .… Multiplier

20     0.8    40T, 50T     10
22     0.727 (recurring)    40T,  55T     10
24     2/3      40T,  60T     10
26     0.61538     40T,  65T     10
28     0.57142 (rec.)     40T,  70T     10
30     0.533 (rec.)      32T, 60T     10
32     0.5      30T, 60T     10
34     0.470588      32T, 68T     10
36     0.44 (rec.)      20T, 45T     10
38     0.42105     32T, 76T     10
40     0.4 20T,      50T     10
48     1/3     20T,  60T     10
60     0.266 (rec.)      20T,  75T     10

The above are simple gear trains, but in two cases, namely 34T and 38T, it will be necessary to work out a compound train in order to achieve the desired result. It is also important to realise that, if a new set of wheels is to be manufactured to suit a worm with a number of starts exceeding ten the blank will have to be "sped up" by the number of starts involved- i.e. for a fifteen start worm, fifteen times. The cutter must also describe an arc defined by the diameter of the largest worm, as the resulting helix angle is dependant on these two factors. I have omitted details for the single start wheels, as if the above can be cut, no details are necessary. The cutting process is one of generation, and is straightforward whilst underway, unless the desired number of teeth is actually divisible by the number of starts on the largest worm- i.e. the one which defines the sweep of the cutter. When this is the case it will be necessary to make up an indexing catchplate in order to get the cutter bit in the right place at the right time. An alternative is to make up a flycutter that will take a number of bits equal to the number of worm starts involved - this will save a great deal of time and cutter wear, obviating the need for the indexing catchplate entirely. 
Pittler Worm and Wheel details:-
Here, as far as the Pittler Company's working practices will allow, is a definitive listing of not only the worms and wheels supplied new with the B2 as standard, but some additional and important information for anyone contemplating cutting a set, as well as the technical information required. It is important to note that the practices of the Pittler Co. were not quite as clinical as modern practice would dictate, though fortunately a great amount of leeway is available to anyone cutting wormwheels in the manner that I have previously described, and the "Invention" works took full advantage of this fact. The helix angle,  p.c.d and diameter of both worms and wheels fluctuated wildly, but as they were produced as matched sets for the machine with which they were supplied, as long as they ran together, this would not have mattered. It is important to note, therefore, that I have only been able to perform my calculations from averaging the characteristics of those sets from which I have been able to acquire data. In this respect, because I have adopted the "clinical" measurements and dimensions that will produce a theoretically perfect set, I I have to point out to anyone trying to complete an extant set that they will very likely run into trouble if they expect to be able to use the following calculations as anything other than a guide.
The Worms
There were three worms supplied new with the machine: a single-start, five-start and a ten-start. However, it was possible to purchase any worm to order and, according to a document written by George Adams (Pittler's U.K. Agent) in 1905, it is clearly stated that worms up to twenty five starts could be supplied, and more importantly, that twenty and twenty five start worms were actually held in stock. If a set of worms and wheels is to be produced from scratch, it is important that the throat radius of the wheels corresponds to dimensions derived from the minor diameter of the largest worm that will be used, and therefore it must be decided whether or not the maker thinks he will require the additional screwcutting and spiralling versatility afforded by such larger worms, and make the wheels accordingly.
Let us start with the single start worm. The material is steel, preferably "machinery steel" i.e. EN3B, EN8M, "twenty carbon", etc. This is the easiest worm to produce in the lathe. The helix angle dictated by the pitch diameter (p.d.) has been kept to a minimal amount, and therefore any correction factor is so small that for our purposes it can be assumed to be negligible. For those who feel the need to go deeper into that statement, the figures I shall describe produce a helix angle of 3 degrees 4 minutes and 40 seconds, and a thread-to-thread pitch error of about one and a half "tenths". Therefore, the screwcutting pitch employed is simply that used for 20 d.p., i.e. 0.1571" per headstock revolution. The changewheel set up is as follows:

20 d.p - 55x40
              35x50

The blank diameter should be 1.030" giving a p.d. of 0.930". It is strongly advised that a thread milling set up is employed, using a No. 1 (rack form) 20 d.p., 14 1/2p.a. cutter, but for those who must use a single-point tool, it should be of a 29 degree (inclusive) angle with a flat on the nose measuring 0.049", and set over, if necessary to correspond with the three-degree helix angle. The depth of cut required (D+f) to produce the full form is 0.108". Please note that I have selected a pressure angle of 14.5 degrees simply because I happen to have the necessary gear cutter of this p.a.. This is what the originals would have been cut as, but there is no reason why a 20 degree p.a. cannot be used instead, with the blank diameters that I have listed here.
The five, ten and greater multi-start worms become more involved, as due to their sizes, they carry a far larger helix angle than can be ignored. I have taken this angle to be 10 degrees 18 minutes and 32 seconds, and this corresponds closely to an example which has been measured. The angle selected allows a simple screwcutting gear train to be employed, with a 32T gear as driver, a 25T as driven. It is important to note that all of the multi-start worms will carry precisely this angle, and it is this that allows them to use the same multi-start wormwheels. The calculation for the pitch correction in the ten-start worm is as follows:

Corrected pitch  =   cosine of the helix angle
                                  uncorrected pitch for 20 d.p.

Then the p.d. can be established by transforming the equation below:

Tangent of the helix angle   =  the lead of the thread      (ten times the linear pitch in this example)
                                                           The p.d. x Pi

It is also important to appreciate that the angle was very carefully arrived at, theoretically, in order to find a simple gear train to produce the necessary linear pitch without having to resort to the method of continuous fractions. This pitch is 0.160", and uses two gears, a 32T and a 25T, as stated earlier. This arrangement however, requires the headstock mandrel to "overdrive" the leadscrew, which would be considered bad practice if the procedure were driven from the headstock. Therefore the lathe must be driven from the quadrant, or, if the thread is being milled, it should be possible to drive from the headstock via a mandrel handle, the load being smaller. As multiple starts are involved it is also necessary to make some form of indexing attachment which will cover the divisions required. I have given thought to the method that I shall use, and as I plan to use a mandrel handle, I am considering the use of my rotary table, secured to the faceplate, for the job. Alternatively, an indexing catchplate will need to be constructed--this being a more elegant solution.

The sizes of the blanks are:
5 start - 1.500"
10 start - 2.900"
20 start - 5.700" (non-standard)
It is also worth noting that two additional worms, of the same dimensions, though slightly longer, were used with the Pittler dividing head in both the original and later, improved form. These were one and five start. The Pittler spherical-turning attachment also required an additional 80T wormwheel and corresponding single-start worm.

The Pittler Wormwheels
To clarify, the full set for the machine as standard comprised seventeen wormwheels, three single start, of 30, 50 and 80 teeth, all throated to suit the single start worm, and the rest of multi starts. These were of 20,22,24,26,28,30,32,34,36,38,40,48,60 and 80 teeth. Additionally, five wheels were made available by the Pittler Co. for use, specifically, with the early-type dividing head. These were of 25, 50 and 100 teeth of multi starts, and 25 and 100 of one start. These were not shown in dividing tables supplied by Geo. Adams, who later produced a superior, drilled, design of head, and their necessity in general dividing is not without question. As with the worms, the single-start wormwheels do not require any correction for the helix angle, and the p.c.d. can be regarded as the same as a 20 d.p. spur gear; but the multi-start wheels must be compensated.

p.c.d.  =  number of teeth     Gives the uncorrected figure
                diametral pitch

Which can then be divided into the cosine of the helix angle. If 0.100" is then added, we arrive at the blank diameter at the centre of the worm-wheel throat. The list for the multi-start wheels is:
20T- 1.1160"
22T- 1.2176"
24T- 1.3192"
26T- 1.4208"
28T- 1.5224"
30T- 1.6240" (N.B. this is the multi start, the single start has a smaller p.c.d.)
32T- 1.7256"
34T- 1.8272"
36T- 1.9288"
38T- 2.0304"
40T- 2.1320"
48T- 2.5384"
60T- 3.1480"
80T- 4.1640" (N.B.- this is the multi start - the single-start 80T has a smaller p.c.d.)

The diameters of the single start worm-wheel blanks can be arrived at by the formula N+2/d.p.
The tool for cutting the teeth should be of the same form as that for the worms, the depth of cut should be 0.108", as per the worms, and it is important for the multi-start wheels that the tip of the tool describes an arc equivalent to the major diameter of the largest worm to be used, plus 0.016". This last correction takes care of the factor "f" incorporated in the cutting tool, which is one-tenth of the thickness of a 20 d.p. tooth across the pitch line, and provides clearance below the dedendum of the tooth form. The single-start wheels will have one worm all to themselves and, using the same 0.016" correction, are to be cut to match the single-start worm of 1.030" diameter. The throat radius to be cut in each of the two types of wormwheel should be equivalent to the radius of the worm in question, less 0.100".

Using the Worms and Wheels for Screwcutting
Its not terribly useful to have a full set of worms and wheels, if their use is not fully understood. Fortunately, it is blissfully simple. For the B2, the proper equation is, for screwcutting-

For any required pitch, the wheel wanted = Pitch (t.p.i.), times worm threads
                                                                                                5

Which, more simply, means that a five start worm will cut a number of threads equal to the number of teeth on the wormwheel with which it is meshed, a ten start will cut half the number, and the single start will cut five times the number. From this it will be seen that the production of a twenty start worm will allow screwcutting from 5 t.p.i. to 10 t.p.i. with half pitches, on top of the standard range. By reversing the positions of the worm and wheel, i.e. by putting the wheel on the