Registration Numbers, Makers' Plates and Design Changes
The decline of the Evans Company in the late 1950's
As with all manufacturers of steam pumps, each pump was given a unique identification number and the Joseph Evans Company was not an exception. Towards the end of the 1950's, J Evans & Sons were experiencing trading difficulties for several reasons, including for example, that electricity was rapidly replacing steam as a source of power and traditional industries using vast quantities of steam were closing down. Orders for new pumps by new customers were minimal and repeat orders from existing customers, along with spares were insufficient to maintain a healthy turnover. (The traditional and existing customers included heavy chemical, mining, gas, coke and tar industries - all now in rapid decline).
In the early 1960's J Evans were bought out, loss making departments closed and the core business of rotating reciprocating steam pumps eventually purchased by G J Weir of Cathcart, Glasgow in 1962. During this period of uncertainty existing customers were supplied from the substantial stock of completed and partially completed Evans pumps and spares. Some of these pumps had already been allocated registration numbers on Evans works plates, whereas others, obviously of Evans manufacture, were given new works plates by Weirs with a new Weir numbering system. Sometimes pumps had two numbers, an original Evans number often hidden under the new Weir works plate. An example of this occurs with Reliable pump, 4&3x4, which has Reg No. 59144 on both the chassis and the cylinder flange, but the former had been covered over with a new Weir plate showing No.322125. For these reasons the chronological order of registration numbers issued during the late 1950's and early 1960's appear to be random and confusing, date-wise, to say the least.
I understand that many of the original Evans records held by Weirs have since been destroyed by fire, however, prior to this unfortunate incident I managed to obtain through Weirs, order details from the original Evans archive for ten pumps confirming registration numbers, type, size, despatch date and original customer; furthermore, my previous employment within the carbonisation division of the National Coal Board enabled me to confirm additional dates/registration numbers for many more Evans pumps used on NCB sites such as coke, smokeless fuel works and tar plants. Thus I feel I have sufficient records to be able to estimate with a reasonable degree of certainty the manufacturing/despatch date of most Evans steam pumps between 1905 and 1958.
Changes in Design and Manufacture
Prior to 1905, there appears to be relatively few preserved examples, although the Evans horizontal "Reliable" pump is clearly recorded in the 1894 Joseph Evans Catalogue. The catalogue clearly shows that these pumps were constructed and assembled on to a common bed plate, with both cylinders and separate bearing pedestals bolted down onto the bed plate. It seems that this method of construction changed during WW1 resulting in a new design of chassis with integral bearing pedestals and vertically machined end faces where each cylinder was bolted onto these faces. This change of chassis design appears to have occurred around registration number 19000 as confirmed by two examples in the collection viz:
- Reliable, No.18949, 5&4x6 constructed on a common bed plate
- Reliable, No.19473, 5&3x6, constructed with the new chassis, the latter with a confirmed date of manufacture of 1916.
Illustrations showing the various methods of construction to follow
Other changes in design seem to have been implemented over a period of time, an example being the method of attaching both piston rods to the kite (or banjo). The early method of attachment was to employ threaded piston rods with a lock nut screwed into both ends of the kite and this is the method employed on pump 18949. On pump 19473, however, the present method of attachment using tapered cottars through slotted piston rods is employed, nevertheless, on an even earlier pump, built 1905 with Reg No. 12011, a "Reliable" Pump of size 8&5x9, has cottar attachments, but still uses a common bed plate and individual bearing pedestals and cylinders. So one must conclude that design changes were implemented over the range as and when stocks dictated rather that on a particular date. Close examination of the pump engravings in the 1894 J Evans Catalogue, 3rd edition, revised to 1904, shows the tapered cottar attachment for attaching piston rods to the kite, but still employing the common bed plate. Therefore in general one should be able to conclude that any Evans pump with a common bed plate type construction was probably made before WW1 and would have had a Registration number less than 19000.
Another clue to date of manufacture applies to the smaller Reliable 4x4 pumps. The initial design of the new chassis with vertical end faces to locate the two cylinders and integral bearing pedestals included an 'open' gap in the casting under the kite and crank for the smaller pumps. Examples of these pumps had registration numbers after 19000 but sometime between 1929 and 1938 the chassis design changed to provide a closed-in casting under the kite and crank. The corresponding registration numbers appear to be between 27000 and 30000 when the chassis changed from an open to a closed-in casting.
Casting Marks and 'E' Numbers
Yet another indication of age may be gained from the casting marks actually cast into each component. The older pumps tend to have the full Evans name cast into such items as steam chest covers, liquid cylinders and on the vertical pumps on the pump bodies or even a decorative name such as "The Ram Pump". Very early pumps have the Evans "Lion" mark cast into the air bottle.
Between wars the castings tended to become plain with possibly the exception of a simple number cast on to the steam cylinder indicating the diameter, however, slide valves, eccentric straps, con rod distant pieces etc still had the respective sizes cast into each item. After WW2, it seems that Evans started to out-source the casting side of their business and to aid identification, 'E' numbers started to appear. These were more commonly found on spare parts such as covers and valves, but complete liquid and steam cylinders have also been found with 'E' numbers. So one can be fairly confident that if parts of an Evans pump have an 'E' number mark on any of the castings these components were of much recent manufacture and probably post WW2. Similarly any steam cylinder without a registration number, with or without an 'E' mark should be regarded as a spare replacement purchased from Evans at a later date (eg frost damaged replacement).
Flywheels
Scanning through all the Evans Catalogues currently available, all flywheels appear to be of six spokes only and of a straight plain design. Small wall mounted and large circulating pumps have split flywheels but there are no indications whatsoever of flywheels with curved spokes being fitted to Evans Pumps. A heavy duty flywheel as an optional extra could be supplied for specific duties, but even these had six plain and straight spokes.
Crankshafts
Early crankshafts were forged in-house and this remained so until after WW1 when crankshafts, particularly for the 4x4 reliable pumps were sourced externally. These may be identified by the manufacturer's mark stamped on the crank arm, examples being C.C.E & Co and JWB; they were machined as opposed to being forged. The replacement new twin circulating pumps supplied to the Caerphilly tar works in 1962 came fitted with machined crankshafts, including the spare as well as Weir numbers!
Works Plates
As well as significant changes to the design of the main chassis/bed plate castings during the early part of the twentieth century, the design of the works plate and its location also changed again giving a good indication to the date of manufacture/despatch. Initially all complete J Evans pumps had a unique serial number stamped somewhere on the castings. The normal place was on the steam cylinder flange and often on the chassis flange too, other common places included the flywheel boss, crank shaft end and air bottles. In addition Evans fitted a separate brass works plate confirming the registration number. It is the changes in the design of these plates that also give an indication to the date of manufacture/despatch. Early examples of brass plates as illustrated appear to be fixed on to the steam cylinder curved surface on the upper side showing number and nominal size stamped into the brass surface. This style continued towards number 20000 probably just post WW1. These plates were relatively thin and along with their fixing location tended to expose them to wear, tear and damage, subsequently they often became detached from the cylinder and lost in time.
From the early 1920's a new makers plate was introduced employing a bespoke thicker cast brass plate each individually made as the registration number was actually cast during manufacture, although the pump dimensions were still hand stamped for each respective pump. The location where each plate was attached varied, especially on the smaller pumps, but in general they were attached to the chassis, often on the steam end flange on larger pumps or the main central area between the cylinders for the smaller and duplex pumps. On vertical wall and floor mounted pumps the plates tended to be fixed on the side just below the steam cylinder. These plates were larger than the pre-WW1 plates and less prone to damage/loss. Towards the middle of WW2, a further change occurred, possibly a cost saving. A standard blank brass plate was introduced, the same size as the cast plates but made from a thicker sheet. These blanks were machine stamped leaving a suitable space for the size and registration number to be added later by hand stamping. This type of maker's plate continued in service until about 1954, when a totally new design was introduced of a larger size and made from aluminium sheet. These plates were pre-printed in black with the individual number and size hand stamped as before. These plates were not very durable and many suffered from corrosion making identification very difficult. The black printing also tended to rub off as well. These plates were the last type to be used and continued in use until 1958. Between 1959 and 1962 the application of works plates was very unclear and unfortunately remaining examples of Evans pumps despatched during this period are few. As previously mentioned some pumps were despatched with original Evans numbers whereas others had new Weir numbers applied while some pumps had both! For example in 1962 a new naphthalene plant was being commissioned at Caerphilly Tar works. Evans pumps had been specified and these were eventually assembled and supplied by Weirs, but were now fitted with brass works plates with new Weir registration numbers of an unknown series, (J 40400/1/2) These pumps were obviously of Evans origin as they still had the Evans name cast into the main castings. Other ex-Evans pumps supplied during 1962 and later had new Weir identification plates employing a long thin brass sheet, preformed and pre-printed with the number and size hand stamped - the Weir number being in the 322xxx series.
Summary of Evans Registration Plates
| Plate Type | Date | Range of Registration Numbers |
| Stamped Brass Strip | to 1917 | up to 20000 |
| Cast Brass Plate | 1918 to 1941 | 20000 to 34000 |
| Stamped Brass Plate | 1942 to 1953 | 34000 to 41000 |
|
Stamped on Printed Aluminium Plate |
1954 to 1958 | 41000 to 70000 |
|
Weir Plate on Printed Brass Plate |
1959 to 1962 | 322xxx series |
Notes: The above table indicates a very high build from 1954 to 1958. Examples of saved Evans pumps with registration numbers in this range appear to be few. One can only assume that either the number series may not be sequential or additional items were included in the system.
In the table below, the following pump dates and registration numbers have been verified through G J Weir and Company by cross reference to the original Evans archive records prior to their loss by fire
| Registration Number | Pump Type | date | location |
| 19473 | Horizontal Reliable 5&3x6 | 1916 | Llanbradach |
| 26900 | Horizontal Reliable 4&2x4 | 1929 | Bedwas |
| 29112 | Horizontal Duplex 5&6x5 | 1934 | Gateshead |
| 30934 | Horizontal Reliable 4&3x4 | 1938 | Caerphilly |
| 31308/12 | Twin Reliable QA 6&5x9 | 1939 | Caerphilly |
| 37673 | Horizontal Reliable 6&6x9 | 1949 | Pontypool |
| 37935 | Vertical Ram 5&4x5 | 1950 | Nantgarw |
| 39275 | Horizontal Duplex 5&4x5 | 1954 | Caerphilly |
| 43141 | Vertical Ram 5&4x5 | 1955 | Aberaman |
However, there are some anomalies that cannot be explained - see below
The following Evans Pumps were ordered by the NCB for its Avenue, Chesterfield and Caerphilly, South Wales tar plants. The Avenue site was for a brand new tar works and the Caerphilly site for a new extension to its existing distillation works. Both sites were under construction in the early 1950's. One can only assume that these pumps were given registration numbers at the time of the initial enquiry rather than the date of actual manufacture.
| Registration Number | Pump Type | date | location |
| 33618A & B | Horizontal Duplex 6&4x6 | 1955 | Caerphilly |
| 33619 | Horizontal Reliable 5&3x6 | 1955 | Caerphilly |
| 33951A, B & C | Twin Reliable QA 6&5x9 | 1955 | Caerphilly |
| 33952A | Twin Reliable QA 6&5x9 | 1955 | Caerphilly |
| 33952B, C & D | Twin Reliable QA 6&5x9 | 1955 | Chesterfield |
All the above pumps were installed between 1953 -1955 at both sites