Pressed Glass

Contents     Site search

Sherriff Family History


Sowerby's Butterfly Pattern


In 1932 Sowerby's Ellison Glassworks launched a new pattern, No. 2552, which included a Ladies Dressing Table set comprising two candlesticks, large and small powder bowls with lids, and a pin dish, all sold with a matching tray.


The design is sometimes referred to as 'Butterflies & Daffodils' but the flowers depicted are without doubt not daffodils, and judging by the shape of the leaves, we believe the flower is possibly a wild rose. The Butterflies, however, are somewhat stylised, but we believe they are meant to depict 'Peacock' butterflies, and that this was chosen to represent the Sowerby 'Peacock' trademark.


The Butterfly pattern dressing table sets were in production for many, many years, right up until at least the 1950's, possibly later. When we started photographing our glassware for inclusion on this site, we noticed that there appear to be several variants of the candlestick, and we have since begun to record the differences.


There are a number of factors that need to be remembered when viewing these items. Firstly, many of the original sets supplied by Sowerby were acid-etched (frosted) with some parts polished, notably the flowers & butterflies. We understand that the factory ceased using the acid-etching process during the Second World War, and although they experimented with sand-blasting with quite some success, there are significant differences between the processes. Although these differences can really only be seen on close inspection, in the case of the Butterfly candlesticks, the shape of the item created a problem.


In the early acid-etched candlesticks, (click on the left-hand picture to see enlarged view) the disturbance of the surface was comparatively minor, and this allowed the flowers and butterflies to be polished as they stand proud of the skirt surface. Sand blasting (right-hand picture) is noticeably more aggressive, however, and it would probably have been difficult to mask off the flowers etc. due to them being raised above the surface. To overcome this, sand-blasting was only used inside the skirt, creating a somewhat similar visual result, although the neck of the candleholder, previously acid-etched, was left in the as-pressed shiny finish.


The second factor that must be remembered is that Sowerby's did not produce any Uranium glass items after the end of the War as their stocks of Uranium were confiscated by the Government. Finally, another point to note is that the mould for the small powder bowl was broken during the War, and not replaced.


Our research so far is mainly supposition based upon these few facts. We have in our possession a number of candlesticks which have been acid-etched, and we know that this process was only used before 1945. One of these is in Uranium glass, and we believe this to be an exceptionally early pressing for a number of reasons. We also have a pair of un-frosted candlesticks that are made from Uranium glass.


The first thing we noticed about all these candlesticks was the size and shape of the recess where the candle is held. This varies from a deepest of 26.5mm deep to a shallowest of under 19mm. This 25% variation seemed quite a lot, but we concluded that the part of the mould that produced this recess could well have been replaced at some time, or was the subject of significant wear.


We found that the recess size does vary from stick to stick, but that there are generally 3 groups - those around 25 to 26 mm deep, another group at 21-22 mm deep, and finally the group around 18-19 mm deep. The sides of the recess also vary, with the deepest being almost vertical, and the shallowest being very tapered, almost to the situation where a candle would not be held very safely.


Having established that this part of the candlestick varied, we began to look for other differences, and have established to our own satisfaction that there are quite a number of variants. In each case, we believe the depth of the candle recess gives us a good clue to when the item was made, with the oldest being the deepest. Combining this with some of the other factors, plus the finishing process used, has enabled us to make a quide for dating these candlesticks, albeit into rather wide-sweeping periods.


To help with this research we have given each variable element a description, and we think of these as follows:

1.    Flower Overlap

2.    Petal Direction

3.    Broken Leaf

4.    Recess

5.    Veins on wing

6.    Butterfly body

7.    Antenna

8.    Ground top


To assist with our research we have tried to photograph as many of these variations as possible.


The candlesticks will be well known to any Sowerby collector. They stand on average  76mm (3") high, have a flared 'skirt' about 100mm (4") diameter, and are decorated with 3 flowers on each half plus they have two butterflies which appear to have just landed with their wings closed that form handles.


Element 1 - Flower Overlap


 Side 1


On each half of the skirt, there are 3 flowers, which we number 1, 2 & 3 starting from the left. There are 2 petals which overlap, one from flower 2, the other from flower 3. On our acid-etched Uranium Glass candlestick, the petal of Flower 2 is on top of the petal from Flower 3. We now refer to this as Overlap A, with Overlap B having the petal on Flower 3 on top. Clearer photos of these elements can be seen by clicking on these links: Photo A  Photo B. Until recently, we had only seen sticks with either Overlap A or Overlap B, but we have just acquired a stick with Overlap A on one side, and Overlap B on the other.


Element 2 - Petal Direction


Side 2Left


This refers to the lowest petal on flower 2. On our frosted Uranium glass stick, this petal points clearly to the right on both sides. On our Flint & Rosina frosted sticks, the petal on side 1 points almost straight down, whilst the petal on side 2 points to the right. In the third type, the petals on both sides point to the left. We refer to the three types as Right, Centre and Left.


Element 3- Broken Leaf



We have only found two candlesticks in our collection that shows this feature - the Uranium Glass, deep recess, acid frosted example, and one of the pair of clear Uranium glass candlesticks. We believe that the frosted Uranium glass example is the oldest in our collection. This stick exhibits a number of differences from all the others. It has a clarity and crispness that is missing from the others. The veins of the leaves, the stamens on the flowers, the wings and body of the butterflies, every detail is crisp and well defined. There is no doubt that this is how the mould was made, but it would appear that this mould was either only in use for a very short time, or that the leaf was re-cut at an early stage.


Element 4 - Candle Recess



As mentioned earlier, the depth of the candle recess varies. In our 3 acid-etched examples the recess is about 25 mm deep, and the bottom of the recess lines up with the tip of the butterfly's rear wing. In the mid-depth recess type, the bottom of the recess is now aligned with the tip of the butterfly's front wing, and the shallowest type has the bottom of the recess well above. The amount of taper in the recess can also bee seen in the above pictures. For our own purposes, we refer to the deepest as Early, the mid-depth as Middle and the shallow depth as Late. Trying to place these in relation to actual dates or events, all our frosted examples have deep recesses. We have a pair of sand-blasted examples which have the middle-depth recess. We therefore think that the deepest recesses were only made pre-1939, that the middle-depth recesses were perhaps 1940's, and that the shallow recess appeared perhaps in the 1950's (or later!)


Element 5 - Veins on Wing



This is very noticeable on the Uranium glass examples, does not appear on any other acid-etched deep recess sample, but is just visible on a pair of mid-depth un-etched sticks. Those sticks also show Overlap A .


Element 6 - Butterfly body


If you look at the picture under Element 5 above, you will note that the body of the butterfly is quite clear, with segments visible on the thorax, and the legs clearly defined. In addition, the butterfly's head is clearly visible. It would appear that as time passed by, the definition of the butterfly's body and head became less clear, until finally it has no head, no segments on its barely-visible body, and no clearly-defined legs. This loss of definition of the butterfly's body lead us on to Element 7, the butterfly's antenna


Element 7 - Antenna



In early candlesticks, the butterflies clearly have antenna, as shown in the first picture above. This is our frosted Uranium sample, and the antenna are about 12.5mm (") long, with the ends about 11mm apart. Other deep-recess acid-etched samples are shown in the second and third pictures and have antenna which are under 11mm long, and only spread out to about 5.75mm wide. On some apparently later-period sticks, the antenna have grown slightly again as shown in picture 4.


The only 'saving grace' is the shallow-recess, Overlap B, Left Direction sticks. They have all lost their antenna, which have been replaced by a new leaf!! This is clearly shown in the last picture. Our stick with the mixed overlap also has Direction Left, and a leaf replacing the antenna. We believe this may be the last version as many of the other features on this stick are very faint.


Element 8 - Ground Top



This has proven quite difficult to photograph, but all of the candlesticks that we believe to be early have the top surface of the candle-holder ground. The first picture is of our frosted Uranium glass example, which has a ground top. The second picture is of a frosted Rosina example, which also has the top surface ground. The third picture is of a Middle recess stick that does not have the top surface ground.




Our research is still continuing, and we will update this page as and when we have other data of interest. There are a few conclusions that we think we can make already. We believe that there were at least 2 moulds in use. Mould One had Overlap A, Petal Direction Right and Wide Antenna. Mould Two had Overlap B, Petal Direction Centre and Narrow Antenna. Both have been seen with Early (deep) recesses. Furthermore, we are convinced that Mould One started life as the Broken Leaf mould and was re-cut at some early point in its life. Our reasoning is based upon another factor which we have not detailed above, but which will eventually appear here, and that is the number of stamens in each flower. Flower One on our Broken Leaf examples has 10 stamens on the broken-leaf side. Later sticks that have both sides with Right direction have one side with 10 stamens in Flower One. To date, we have not found a stick with Centre or Left Direction that has more than 8 stamens in Flower One.


Mould One is seen with Middle depth recesses, which we believe to be mid- to late-1940's at the earliest, and we have examples of both Mould One and  Mould Two with sand-blasted insides which was only done post-1945.


Either one or both of these moulds were then re-cut to the Left Direction, perhaps in the 1950's. At about the same time, Late Recess appeared. Much of this is guesswork, but it certainly makes sense to us. There are other differences apparent between the various candlesticks, which we also plan to photograph and publish in due course.




If you haven't already done so, please sign our guest-book before you leave! Click here


Back to top of page


Sherriff Family History, the Parish Registers of Modbury & Yealmpton, Devon and other items of Interest

2002-6 Paul and Christine Sherriff Last Updated 17 Feb 2010 10:22

 Hosted by