Letter to President of SA Croquet Association - 8th August 1989.
Editor's note: The Barlow Balls at the time would be XTs or Cs

This is to apprise you of the fact that I have just completed a series of tests on a set of first colour Barlow balls. These balls are the latest Barlow balls to date, viz. 5th August 1989 manufacture, and I was requested to do these tests for submission to SACA for approval of the balls.

The specification chosen is the latest one, not yet legally introduced I believe, submitted to the World Federation recently.

Tom Barlow has chosen the middle to low value of 33" bounce for these balls at an ambient 20° C, which is about one inch below the middle of the specification limits. I believe he has done this to accord with the general wish of players to have balls which more closely behave as the present Jaques balls for the roquet and croquet strokes.

Regarding your comment on spalling, not to be confused with scratch marks or other abrasive damage:- Spalling indicates incorrect moulding time/temperature relationship normally, and was very evident on the early hand mixed/moulded Barlow balls. At a later date Tom moved to automated moulding, with corrected processing. At S.W.C.C. we have played regularly for nearly a year with four sets of balls of this same formulation and these show no signs of spalling at all. I examined them with a magnifying glass at the Club this afternoon (Sunday 6th). Regarding general appearance, Tom seems to have managed to remove mould sprue marks formed at mould intersections, so that the balls look better. Also because there is no final lacquering of the surface, it appears as a spherical surface with milling grooves across it as opposed to the rather rounded effect which lacquering normally applies to each of the small lands between milling grooves on the Jaques balls. This is a permanent advantage of moulding and contributes to the constant diameter of the balls from first use onwards. It is noticeable that early flattening of the lacquered land surfaces leads to an early reduction of diameter on lacquered balls.

The balls are as yet unmarked and 1 feel that now that we have a usable acceptable ball it should be distinguished from all previous Barlow balls, good or bad, and that any future changes should also be clearly marked. Therefore the test figures which follow will have to refer to "unmarked balls handed to me on 5th August 1989."

I understand from Tom that there is no problem here and provision for a logo has been made but not yet implemented.

One more comment on the figures which follow:-

I always presume that standard test conditions should be sea level and 20°C. It should be borne in mind that resilience of most materials varies with temperature, so that the results of tests conducted way outside normal expected playing conditions can not be correlated with standard test results. In the case of many hard compounds such as those used for billiard, croquet or bowling balls the resilience will rise with lower temperatures.

The percentage variation is not very marked with most substances and in any case can be mostly compensated for by the addition of other materials well known to plastics and polymer experts.

The test results are as follows:

Specification for Size:

92.87 mm dia. max. - 91.28 mm dia. min. Variation max. to min. of any one ball less than 0.79 mm, and variation in any set limited to 1.19 mm.

Results

All the balls had a 91.5 mm minimum with a 92.1 mm maximum, and are therefore well within both limits of variation as well as the specification.

Specification for Mass

New value 467.89 gm. max. 439.55 gm. min. Old value 460.80 gm .max. 446.64 gm. min.

Results

Black 456.2 gm. Blue 455.9 gm. Red 455.8 gm. Yellow 45.3 gm.

Specification for Bounce

Specification 31" - 37", variation per set 2" allowed.

Results

The balls were each bounced three times on each of the 4 poles and 2 nodes. These figures are not all given because of unnecessary space but the average values and variations are given as follows:-

As can be seen, the variation per ball did not exceed 1.25" nor did the set variation exceed 0.59" on averaged values. On individual pole/node values the maximum set variation was between 33.60" and 3.20" or 1.4" which is still well within the allowed 2".

In passing, perhaps the revised specification should also-contain the proviso for ball bounce variation in resilience at poles and nodes, as it is a tremendous test of consistency of compounding of any one ball.

The random bouncing of balls can lead to false figures, missing what may be "hard" or "soft" spots on the ball which show up if the ball is bounced on all six points accurately.

I hope that these test results and comments prove useful in promoting this new ball, which seems at last to be fulfilling all our expectations and certainly my own predictions. Perhaps there will be a dozen such on the croquet scene soon, but at least we might have a ball in South Africa suited to our impoverished pockets, and of a technically superior quality.

Kindest regards,

R.B.C. le Maitre ©1989