OxfordCroquet Logo

Dr Ian Plummer

Donation Button

Report on the Barlow Croquet Balls

As produced in the latter part of 1988 (Aug-Sep)
Editor: These are early Barlow Balls, not the current GT series


These latest Barlow balls, some thirty-two in all, were tested during September on two separate occasions for size, mass and bounce, according to the Association specifications and procedures.

At the outset, it should be mentioned that great advances have been made in the manufacture of the Barlow balls, which now utilise the most modern methods of uniform, air-free casting known. In addition, the balls are now moulded in their entirety in a single operation, which explains their consistent results. In particular the milling, being also now moulded, is machine accurate and approximately twice as deep as that of the Walker ball, while retaining a sharper re-entrant angle of ±90°. This is an excellent achievement and for those who understand moulding, a tribute to the design and construction of the moulds.

The other advantages of the Barlow ball composition, such as wear resistance and accordingly ball size retention, homogeneity of construction including pigmentation have been mentioned before. Concerning the composition of the Barlow ball, it is a well-known fact that the properties of resilience or restitution, together with mass or density of such materials, can be modified to suit almost any requirement. It is therefore of passing interest that Mr Barlow has chosen the figures for bounce that he has. This value is almost in the middle of the specification allowance by the Association. To those of a more conservative nature who wish only to play with the ball resilience that they are used to, this value may be somewhat high. However, there is no prima facie reason to stick to the lower values specified.

In an earlier report, 4 Jaques Eclipse Premium balls were tested in all respects, together with 4 original slightly-used Brento Tournament balls. In that series of tests, the Jaques balls gave a corrected maximum bounce of 31.5 inches, the Brento balls 32.5 inches. These values are only just acceptable in the laid down specification of 30-45 inches. Given a choice, the bounce region should be chosen roughly half way in the scale, and this may have influenced the choice of Mr T Barlow. The Barlow formula can be modified to suit higher as well as lower bounce figures, and the real criterion of quality is consistency, not absolute value of test results, as long as these lie within the specified range.

Test Results

Due to the extraordinary consistency of test results on the Barlow balls, it is not proposed to tabulate them, but to give the average values obtained together with their maxima and minima.


The balls were measured across six axial positions. The variation of all these measurements to one place was from 91.5 to 91.8 mm. or 0.3 mm. This compares with the 4 Jaques Eclipse Premium balls, which varied from 91.5 to 92.1 mm. or 0.6 mm. Specification 91.2 - 92.8 mm. The maximum variation in sphericity on any one Barlow ball was 0.2 mm. as compared with the Jaques 0.4 mm. Specification 1.6 mm.


The average mass of the Barlow balls was 458.2 grams, with a variation from 454.8 to 461.0, equal to 6.2 grams. Based on the specified mean of 454 grams and the allowance of ±7 grams, the above values are in the upper half of the tolerance.

The Jaques Eclipse balls also gave excellent mass results, ranging from 453.5 to 456.5 grams. Although only four Eclipse balls were available for measurement the mass variation of only 3 grams was excellent, as mentioned in the previous report, and represented a great improvement on previous Jaques Eclipse balls, as well as being in the middle of the specified value.


The 4 Jaques Eclipse balls were bounced on all four poles and the two nodes, three times each, a total of 72 bounces. The average corrected bounce obtained was 31.2 inches, with a maximum of 32.5 inches and a minimum of 29 inches, a variation of 3.5 inches.

The Barlow balls showed the most remarkable consistency in bounce that 1 have yet seen since I began testing Croquet balls. The regularity of the bounce was monotonous at 38 to 38.5 inches. By going to a further split of 1 inch divisions, it is possible to include 37.75 and 38.75 in the range, giving a variation of one inch about an average of 38.25 inches.

It is only fair to state at this point that the Jaques balls tested were the Eclipse Premium, not the Eclipse Premium Selected. Possibly the bounce values of the Selected balls would be closer than the 3.5 inches of the Premium, but it is also fair to say the Barlow balls need no selection for any particular characteristic, since their consistency does not seem to vary.

In a letter from Professor Bernard Neal to Mr C Jaques, copied to me, dated 26th August 1988, Professor Neal quotes the specification for bounce as being between 30 and 45 inches dated from 1984, the only specification actually accepted and adopted by the Laws Committee. The Barlow balls therefore lie just above the mid-point of this specification.


The advent of this latest double axis spun cast Barlow ball heralds a new standard of precision in Croquet equipment manufacture and testing them has been a pleasurable experience. Croquet will benefit in the same way that many other sports have done when their equipment is manufactured to exacting standards of consistency.

R le Maitre

Author: R le Maitre
All rights reserved © 1988

Updated 28.i.16
About, Feedback
on www.oxfordcroquet.com
Hits: 10044