The results given for the tests have been averaged.

Strike Shot

Figure 1 shows the results of the strike tests. In each test a single ball was struck by a mallet falling from a given height and moving through a set arc or angle. The mallet was held suspended in a frame, pivoted about the end of the handle, and released from the same height each time. The impact or strike test was conducted on a bowling green, and the balls were struck from a set point on a board, the base of the frame. From here they moved out onto the bowling green. A sketch of the apparatus is attached as Figure 2. The energy available in each case was that of a mass of 850gm. falling through 583mm. Neglecting the very small windage and friction losses. This energy amounts to 49500cm. grams.

Figure 1. Strike test on one ball across court. The graphs are of the averaged results

1. The initial velocity of the Walker balls was 3.60 metres/sec., and by calculation based on the graphs of Figure 1 the velocity of the Barlow balls was 3.42 m/sec., of the old Jaques it was 3.27 m/sec., and 3.13 m/sec. for the new Jaques balls. [Editor: at this time there would have only been Barlow 'C's or their predecessors]. Thus the energy pickup of the Walker balls was about 29990 cm. grams or 60% of the available mallet energy. This figure is reduced to 55% for the Barlow balls, 50% for the old Jaques and 46% for the new Jaques balls. These figures are to be compared to those for the concreted rigid steel plates, which were 69%, 61%, 45% and 42% respectively. The difference is attributable to the improved energy pick up from an elastically deformable object such as a mallet. Even so, there was still a proportion of the energy left in the mallet since the swing continued for a short arc after impact in each case. This result is clear evidence of the different strike power which would be required if, say, rubberised mallet ends were to be allowed.
   
   Figure 2. Croquet/Roquet Test Rig
2. According to the strike test results of total distance rolled, then if a Walker ball is struck hard enough to travel the 33 length of a court, a Barlow ball will travel 31.4 yards with the same blow, an old Jaques ball 29.9 yards and a new Jaques ball 28.7 yards.

Croquet Stop Shot

In this test, the same mallet energy was applied as in the strike test. The total distances in metres run by the balls in each test were averaged, and the results for both the forward and backward balls are as follows:

1. Type Of Ball	Forward	Backward	Total
   Walker	8.285	1.975	10.260
   Barlow	7.085	1.910	8.995
   Old Jaques	6.830	2.105	8.935
   New Jaques	6.230	2.327	8.557

As suggested, the total distance run by both balls is very close to the maximum distance averaged by the same types of ball in the croquet test.

Walker compares 10.260 to 9.830 + 4.4%
Barlow compares 8.995 to 8.340 + 7.8%
Old Jaques compares 8.935 to 8.950 -0.2%
New Jaques compares 8.557 to 8.550, ±0%
 

3. These figures, given the same elastic blow of the mallet, indicate firstly that the court resistance due to ground and grass is mostly of the form of work done, and therefore roughly proportional to the distance run and not either velocity or velocity squared (as for instance air resistance would be).
4. There are major differences between the balls when we consider the forward and backward balls individually. Here it is plain from the tests, as has been found in real life practice, that as elasticity increases, so does the disparity in the croquet stroke between forward and backward balls. Calculation of the ratio of distance travelled by the forward ball to the backward ball gives a value of 4.2 for Walker, 3.7 for Barlow, 3.2 for old Jaques and 2.7 for new Jaques balls. It is clear from these figures that it is more difficult to get the back ball up and prevent the forward ball going too far with the more elastic balls.

Croquet Roll or Drive Shot

Here, more mallet energy is imparted because of the follow through applied.

In the case of the tests in this report, the results were achieved by weighting the mallet head with a further 1360 grams, making a total of 2210 grams.
 

5. From these results it appears that the Barlow ball plays similarly to a used Jaques ball on the drive or roll shot while the Walker ball is still difficult to control.

Roquet Shot

The results were as follows for the forward ball only:

6. The roquet test, performed with same mallet impact as in the other tests apart from the roll, shows that once again the behaviour of Barlow and old Jaques balls is similar, while the Walker balls continue to roquet 20% more than the old Jaques/Barlow average and the new Jaques roquet to a slightly less distance.
7. The reason for the new Jaques balls roqueting less, and noticeably so on the longer distances, may be due to the impact between balls being partly cushioned by what might be described as a "meshing" effect between the milling in the outer covering rather than a compressive effect directly on the core. Be that as it may, the author has found on occasion that in a hard roquet the tendency is for the striker's ball to lift slightly on impact, and for the roqueted distance to be greatly reduced. This effect could be the result of entirely different causes, such as the effect that with a rougher milling, balls will travel slightly higher above grass level because of increased surface friction. This would occur especially at higher speeds and would account for an above centre line impact between striker ball and roqueted ball, with the result described. This effect of surface friction, namely the travelling height of the ball centre above grass average level, is reduced as the ball begins to roll rather than slide, and so is reduced in the case of a longer shot to the roqueted ball.
   
   In passing, it should be noted that no part of the surface of a new Jaques ball is spherical, because the actual surface is broken up by the milling into a number of rounded points or ridges. This is corrected in the Walker and Barlow balls, which offer a spherical surface at all positions, only slightly interrupted by the milling.

Conclusions

1. If the specification in the new rule book of 1986 is to be correctly applied, then the Walker and Barlow balls are closest, while no Jaques ball, new or used, passes. This is indeed a strange circumstance.
2. Of the two types of modern ball, homogeneous Walker and Barlow, both could be modified to a small degree to achieve specification requirements.
3. In the author's opinion, the Walker balls are too elastic to be controllable. The pass roll becomes a difficult stroke to perform without a scrape or other fault, and the roquet too easy, requiring no heavy blow.
4. For the Jaques ball to pass specification, much work is needed in the preparation of the core to cut down unacceptable mass variation and improve bounce. This may not be possible with the known variations in timber which occur, which factor does not affect the other two balls.
5. Considering the ultimate toughness requirements of a croquet ball vis-à-vis high speed contact with a hoop wire or other rigid object, the construction materials for a modern croquet ball will probably include nylon, polyester resin and polycarbonate resin, filled in each case with glass or carbon fibre and additive fillers such as pigment and light and heavy fillers to correct bounce and specific gravity.