Article 5. Written September. Copyright Gavin Bottrell. 

Contact me using info@timewarpgolf.com

Can the Proportions of Feathery Golf Balls indicate how old they are? 

Previously I have written about the roundness, weight, and density of feathery golf balls. This article is concerned with the proportions of the leather pieces used to make the case of the ball and proposes a new theory regarding their manufacture.   

Gowland¹ states that feather balls made in the 18^th century were larger and had narrower central strips and larger end caps than those made in the first half of the 19^th century. This idea is worth investigating further as it may assist the dating of very early extant balls. The following illustrations first appeared in 1896 in Golf Illustrated², before being republished in 1978 by Henderson and Stirk³. 

Figure 1. Drawn illustrations of Feather Ball produced in 1896.

The ball described as being very old is of larger size and more oblong, however, it does not have a narrow central strip. As shown in my previous investigations feather balls were made to be as round as possible, and some well-used examples have retained a high degree of roundness. A very small number of balls have been sold at auction over the last few decades with an attribution of dating from the 18^th century. These have been larger in size and the sewn seams appear cruder than those on named balls we know were made in the 1840s. The surface texture of the leather also appears rougher with surface fissures, much like the illustrated ball. We can say with certainty that large sized 34 pennyweight balls were being made right up until introduction of the gutta ball in 1848, so size by itself is not an indication of great age. Comparing the relative crudity of two objects is a recognised method of determining which is likely to be older in archaeological terms. There is evidence that feather balls were used from the early 17^th century, and in my opinion likely before that, but no investigations have been carried out to determine whether the feathery underwent development stages as later gutta and wound balls surely did.

In 2008 during a visit to Royal Blackheath to provide hickory clubs for their quadricentennial celebrations I was very excited to be able to view their historical artifacts. I was told that the early silver balls attached to the silver trophy club of 1766 were real balls coated with silver, and in 2012 during a tour of the R&A clubhouse at St.Andrews I was told the same thing. Over the last decade I have often thought about whether this information could be true. I started to wonder whether any trends are evident, in terms of size and proportions, over the century or so that they were being attached. Over the last two years I have revisited Blackheath and St.Andrews and was able to photograph the silver balls; a dated selection is shown below.

 

Figure 2. Silver Balls attached to the R&A original silver club dating between 1754 and 1764 - wavy seams and prominent cross-stitches.

Figure 3. Silver Balls attached to the R&A original silver club dating between 1771 and 1805. Straight seams and many have no visible stitches.

 

Figure 4. R&A Silver Ball of 1812  – wavy but no cross-stitches.

  

Figure 5. R&A Silver Ball of 1825 – straight seams with lots of tiny stitch marks.

 

Figure 6. Silver Balls attached to the Royal Blackheath original silver club dating between 1769 and 1787 – sine-wave seam lines enhanced in black for clarity on some.

Figure 7. Silver Balls attached to the Royal Blackheath original silver club dating between 1789 and 1840. Sine-wave seams, no seams, cross-stitches and U-shaped stitches.

Upon detailed examination it is evident that some of the balls are dented – no doubt the result of some mishandling. This is the first clue as to whether real balls are contained within. Whilst it is necessary that real balls compress when struck with a club, they are very firm and cannot be compressed when squeezed in the hand - it is impossible to put small dents in them. A wide variety of small objects have been partially or fully encased in silver for a multitude of reasons. For it to be successful the object needs to be dimensionally stable at room temperature otherwise small cracks are likely to appear over time, particularly around areas of engraving, or where hanging fixtures have been attached. I did not see any such cracks on any of the silver balls. There are no large gaps in any of the seams and the stitch marks are very uniform – this is at odds with what is often seen on real balls. These observations clearly indicate that the silver balls are hollow spheres and that real balls are not contained within. Whilst this was a disappointing realisation I began to notice some interesting details about the silver balls.    

A silversmith commissioned to make a silver ball would’ve had to decide what method of manufacture to use. Hollow spheres can be produced by different casting techniques or by beating flat pieces of silver into three dimensional forms. Surface features can be depicted either by incorporating them into the casting or by engraving them directly onto the silver object after molding. The seams and stitches on most of the silver balls appear to have been cast in, though some were definitely engraved post molding. Most of them are uniformly spherical, and the lack of localised bulging, particularly around the closing hole, suggests that impressions were not taken from real balls. On real balls the closing stitches sit slightly proud of the surface yet this is not present on any of the silver balls. On some balls the closing stitches are indicated by finely engraved loops (see Figure 5). The earlier dated balls tend to have wider and deeper seams and stitches, and some are not very spherical, such as the 1771 ball of Ninian Imrie. This might be why people have believed they are plated real balls.  

All of the silver balls conform to the same pattern namely a central strip and two end caps. I have never seen an authentic feather golf ball that isn’t constructed like this, and this is evidence that it was the accepted pattern from the mid-18^th century onwards. Whilst this might seem unimportant, or obvious, it is worth considering that the leather casing for other balls, such as those for cricket, real tennis and stowball, were cut to a different pattern.  Tennis balls were made up of four pointed ellipses which meet at the north and south poles, and a handful of ancient ones dating from Tudor and Stuart times have survived⁴. From my ball-making experiences I know a nicely shaped ball can be made using this four-quadrant pattern, however, the poles are weak points and this can be observed on some ancient tennis balls where the leather has peeled back. This is why cricket balls are constructed with the two hemispheres offset at ninety degrees. An alternative method of constructing a leather sphere is to sew together two intersecting figure-of-eight pieces as done on modern baseballs. Modern tennis balls still have seams of this pattern molded onto their surface. These seams exaggerate spin put on by the player or pitcher and this is an intrinsic part of playing these games. Of course, golfers mostly desire their ball to fly relatively straight, and is likely why the figure-of-eight configuration has never been used. Interestingly, in 1981 a football was found in Stirling Castle⁵ believed to date from before 1540.  This ball measures 15cm in diameter, roughly half that of a modern football, and has an outer sewn case and an inner pig’s bladder which would have been filled with air. The ball looks like a huge feathery as it is constructed of a central strip and two end caps, and it is very possible that the craftsman who made it was familiar with both footballs and golf balls as both games were banned by King James II of Scotland in 1457. The game of football had been banned in an earlier act, but it is not possible to say which of the games is older and whether the design of footballs influenced the design of golf balls, or vice-versa. In the early 19^th century football was still being played at Harrow school with a leather ball shaped the same as a feathery golf ball - these were termed “pork pie” balls⁶, before the familiar shaped multi-panel balls were introduced in the latter Victorian era.

The fact that the silver balls are so detailed indicates the makers must have seen real balls – imagine the challenge of commissioning one without showing the silversmith a real one. I do not know if any records still exist that reveal who made the first silver balls (and clubs) for the early golfing societies but even if they are known, or discovered in future, I highly doubt they will include any written instructions on the detailing. The most logical and simplest approach would’ve been to supply real items to the silversmith with the instructions “copy this and engrave it with this emblem and these words”. Just imagine the scene as the wealthy gentlemen cast their critical eyes over these very expensive specially commissioned silver works of art for the first time. As with similar pieces produced today a very small amount of artistic is quite likely but I don’t believe the inclusion of details that didn’t exist, or the omission of some details that did, at the whim of the craftsman would have been acceptable. The silversmith certainly wouldn’t want to risk not getting paid or have to re-make an unrepresentative item at his own expense! If I was making a silver ball, let alone a silver club, I would strive to make it as exact a copy as my skills and the materials I was working with would allow.  Therefore, I have high confidence that the silver balls, and silver clubs for that matter, are close representations of the contemporary equipment in use. But before continuing it has to be acknowledged that one ball has stitching that cannot be accurate – this is the 1825 ball of Sir Anstruther as shown in Figure 5. It has a multitude of small lines engraved perpendicular to the seams. If these are to represent stitches there are six times as many per inch than on actual balls! It is possible they were engraved to represent the leather dipping down either side of the seams, similar to how an artist working in pen might portray a sloping surface. This is the only ball I observed out of all the silver featheries attached to the R&A silver clubs that was engraved like this. All the others had stitches spaced at intervals that generally align with real balls.

By comparing the silver balls down the years we can determine what, if anything, changed. In terms of physical size no trend is evident. In terms of the relative size of the central strip to the end caps no trend is evident. Balls that have been attributed as being very old are on the large side and some have thin central strips, but it might be that other smaller balls without thin central strips are as old.  A private collector in the USA, owns a very large ball and he attributes it to being from the mid-17^th Century. It has straight seams and invisible stitches. He kindly advised me that it measures 6.4 cm in diameter and weighs 85g. This is thirty percent bigger in size, and twice the weight, of an average feather ball putting it firmly “off the scale” and it makes me wonder whether this ball was made for golf. It if was, then the range of a good golf shot in the mid-17^th century must have been considerably less than one played in the mid-18^th century. It would also be interesting to practically investigate how larger balls such as this perform when using ancient equipment of the form of the Troon Clubs – I intend to do this at some time.  

Royal Blackheath owned a feather ball that was labelled as being from 1718, although this date is unverifiable. This ball is unremarkable in terms of its diameter and proportions; it is very similar to balls of the 1840s. If the date is correct, then it suggests that the maximum range of golf shots did not change from the early 18^th century right up until the feather ball was superseded by the gutta percha ball.

The R&A silver ball dated 1805 for David Dewar does have a thinner central strip (see Figure 3). This is either because the silversmith decided to engrave it that way using some artistic licence or that balls with thin central strips were being produced at that time. From practical experimentation I know that the dimensions of the three pieces of leather can be altered, and providing they balance each other, an effectively round sphere can be produced. It is surprising to me that there is not more variation in proportions in all the real balls and silver balls I have examined. When I first started looking at pictures of feather balls in reference books I formed the impression that balls made by the Robertsons were smaller, had thin central strips and were “eggs”, and that Gourlay balls were larger, had thicker central strips and were “doughnuts”. The data I have collated shows that this was an incorrect understanding and that both families produced balls of all sizes with no consistent trend in terms of leather proportions. However, it is quite possible that some lesser-known makers, now lost to history, produced balls of quirky proportions and weights.  From my own experience I know that ball-making requires both knowledge and skill and I have been slowly acquiring both by trial and error over many years – starting from scratch I have made plenty of mistakes and have obsessed over some factors which I now know are red herrings. Generations of ballmakers didn’t have to go through this frustrating process – they were apprenticed to a master who taught them all they needed to know – and thereafter appear to have stuck with what they had been taught. But, of course, someone initially had to figure out what would work through trial and error.  Studying the silver balls suggests that a major development in how feather balls were sewn did take place in the late 18^th century.  

The stitching of golf balls

Through my ball making experiments I have learnt a great deal about the stitching of feather balls, and through practice have developed some expertise. The technique used on all extant feather balls is tunnel-stitching, sometimes called butt-stitching. This pulls the edges of the leather pieces together in a relatively straight line save some minor puckering and the stitches are largely invisible when it is done well. What struck me when I first inspected the silver trophy clubs at Blackheath in 2008 was the unusually curved “sine-wave” and “continuous U” seams on the oldest silver balls and the prominent stich marks which cross the seams at right angles. (See Figures 6 and 7).  If the silversmith was accurately depicting the seams on actual balls then it appears an alternative stitching technique was used. The seams on the very early R&A balls (See Figure 2) are not sine-waves, though they have some degree of waviness. However, the seams on the 1812 R&A silver ball (See Figure 4) are of sine-wave form though not as exaggerated as on the Blackheath balls. The trends can be summarised as follows:

1754 – 1770        R&A	Irregular wavy seams with prominent offset cross-stitch marks
1771 Onwards    R&A	Straight seams, no stitch marks
Notable Exception R&A	1812 ball has distinct sine-wave seams

Figure 8. R&A Silver Balls and Seam Shape

1766 – 1790 RBGC	Mix of sine-wave and u-shaped
1791 – 1802 RBGC	Some balls have no seams but have legible dates and motifs so were made this way.

1805 Onwards RBGC	Mix of wavy and straight seams tending to straight in later years.
Notable Exception RBGC	1840 ball has very large u-shaped seams

 Figure 9. Royal Blackheath Silver Balls and Seam Shape

Since starting this article I have obtained a few, but not many, photographs of silver balls attached to two silver clubs belonging to the Honourable Company of Edinburgh Golfers. On the first club dating from 1744 is a silver ball dated 1752 and the sine-wave seams are identical to those seen on the pre-1771 R&A balls; they are wavy with prominent offset stitches. Silver balls attached to the second club dating from 1811 have much straighter seams, with visible cross-stitch marks, but non appear as straight as seen on the R&A clubs post-1771. The amplitude of the sine-wave reduces over the years but is still present, complete with cross-stitches, up until the last silver feathery was attached in 1866.  This date is very intriguing because it is a considerable number of years after gutta percha balls were introduced for play in 1848 and I think it highly unlikely that all the captains between 1848 and 1866 played their golf with feather balls.

All authentic feather balls I have inspected over the last two decades have straight seams, meaning the leather caps and strips were cut without any local deviations. On some balls very small elliptical gaps have opened up between the stitches due to leather shrinkage. These gaps are not enough to be mistaken for a seam of sine-wave form. There are two possibilities: either the silversmiths invented this sine-wave seam shape for decorative purposes or real balls were produced with them. Figure 8 shows one of the early attempts at making a ball using an externally sewn whip-stitch and a piece of rawhide I recently sewed using a running-stitch. 

 

Figure 10. Replica Ball No.8 (external whip-stitch) and rawhide (running-stitch).

When the thread of the running-stitch is pulled tight a distinct sine-wave seam is produced – this does not happen when an external whip-stitch or internal butt-stitch is used. The benefit of the running stitch is that it is much quicker to do, however, it tends to open-up more as the hide dries and is not as strong as only one thread, rather than two, is passing through each set of holes.  

The seams and prominent cross-stitches seen on replica ball No.8 are very similar to those seen on many of the silver balls, although no distinctive “continuous-U” shape is evident. As this whip-stitch can be done externally it is quicker than using the internal butt-stitch yet just as strong. 

Definite trends can be seen on the silver feather balls, and there must be a reason for this. If they are accurate representations of actual balls in use then it appears the ballmakers of St.Andrews had ceased to produce balls using external whip-stitching by 1754 and had adopted the “invisible” butt-stitch from about 1770 onwards. There are a small number of silver balls at Blackheath dating between 1787 and 1802 which appear to have no seams yet carry legible names, dates and motifs. This may reflect the introduction of invisible seams or might just be the whim of the makers. This theory of the stitching techniques changing has not been proposed before and it relies on the silversmiths accurately reproducing what they saw. It seems logical that members would use a favoured silversmith for the work. If so, then it seems logical that a craftsman would simply re-use the mold he’d created previously rather than go to the trouble of making a new one each year. However, we can see small but clear differences in detail from year-to-year which suggests these balls were not batch produced or made from the same mold. Given the differences in detail it appears that a number of silversmiths were commissioned with the work, and hence it becomes even more unlikely they would all include such things as a sine-wave or very wavy seam with offset cross-stitches if they didn’t actually exist. This supports my notion that the silver balls are representations of actual balls in use, but equally it should be recognised that a very small number of balls include details that are clearly not accurate! An example is the number of stitches on the R&A ball of 1825 (Figure 5) – the vertical seam shows seventeen ultra-small stitches, if that is what the silversmith aimed them to be, whilst a real ball has three. More evidence for a change in stitching techniques might be found by analysing whether the price of balls changed significantly around these years in these localities; one would expect the price of internally butt-stitched balls to be significantly more due to the increase in labour time. Any pricing study would need to consider that ballmakers also sold repaired balls as I provided evidence for in my last article.  

What is conspicuous by its absence is that I do not know of an old feather ball with pronounced sine-wave seams or one with wavy seams with offset cross-stitches! The seams on some real balls can appear to be slightly wavy when viewed from a particular angle. The pictures below are all the same ball made by John Gourlay in the 1840s. The seam of the top cap appears relatively straight from one angle (side view) yet wavy in another (top view). This optical illusion is due to the leather shrinking between the stitches.  Of all the authentic  balls I have inspected this one has the most wavy seam, and it is only noticeable when viewed from a particular direction. The amplitude is only a small fraction of that depicted on some of the silver balls and so I think it highly unlikely this is what the silversmiths intended to portray.  

 

Figure 11. Three views of a c1840s John Gourlay ball.  

Conclusions

Either real balls were sewn in such a way as to produce sine-wave or irregular wavy seams with prominent cross-stitches, often offset at the top and bottom of each wave, or this is purely artistic licence on the part of the silversmiths. If a golfing society had patronised a particular local smith to produce silver balls over a period of a decade or more, and that craftsman had exercised some artistic licence when detailing the seams which was acceptable to the incoming Captain, this could explain the major trends in seam shape. However, if only one man had been used it would seem both logical and likely that he would’ve re-used the mold he’d used previously, or batch produced them. And yet, every silver ball I have inspected appears to be unique (ignoring the obvious difference in the engraved name and date); this alone is fascinating. There may be reasons why molds of the time, likely clay, could not be re-used, but as a professional engineer who once designed molded metal and plastic parts, I cannot think of any except personal choice! Some of the balls are clearly of worse quality, with more small surface occlusions than their nearest affixed neighbours, indicating the purity of the metal varies on each ball. Their uniqueness and variable quality strongly suggests that different silversmiths were employed from year-to-year in their production. This leads to the question – how likely is it that different silversmiths, in different geographical locations, decided to exercise their artistic licence in the same way when portraying the seams and stitches? I think it highly unlikely.

Some years ago I inspected a ball with prominent external cross-stitches that was being offered for sale at auction; it only took me a few seconds to conclude it was a reproduction. Some reference books include tantalising pictures of balls with prominent external cross-stitches but I have not been able to track them down and inspect them; they may be authentic but I have my doubts. Very few clubs have survived that can be reliably dated to having been made before 1800 and in my opinion many ancient specimens have been incorrectly dated, in some cases by as much as a century. I am not aware of any extant authentic feather balls that have either sine-wave seams or distinct wavy seams with offset cross-stitches as portrayed on many of the 18^th century silver balls. This could suggest that my theory regarding sewing techniques is a flight of fancy and that the silversmiths employed some serious artistic licence, or it could be that the oldest feather balls extant are not as old as they proport to be. Should anyone know of a feather ball with sine-wave seams please do get in touch – it might be the oldest golf ball on the planet.

Gavin Bottrell