1879.] Spontaneous Segmentation of a Liquid Annulus. 57 



eliminated by many observations. As the observations were made 

 before the theoretical number had been calculated, the judgment in 

 estimating doubtful impressions could not have been biassed more than 

 by the inclination to recognise a constantly recurring number of 

 drops. 



The re-fusion of the drops, owing to contraction, makes it neces- 

 sary to take the impression before the annulus has fallen far, if we 

 wish to ascertain the maximum number of drops into which it has 

 originally divided. In the case of the smaller annulus, 9 centims., 

 and in that of the larger, 12 centims., was the least depth of fall that 

 the apparatus permitted, i.e., that was consistent with getting the 

 disk out of the way. The weight of mercury in the first annulus was 

 26*525 grams, and in the second, 53'05 grams. Assuming that the an- 

 nulus at once becomes cylindrical on its release, the thickness (d) 

 would be *3844 centim., and *4599 centim. in each case respectively, 

 and if the segmentations are determined by the initial dimensions 



(before contraction) according to the law n = —, the number of drops 



7rd 



in the first case will be 13*91, and in the second case (ft 2 ), 16*25. 

 It will be seen by reference to the smallest height of fall in each case 

 that the corresponding experimental values are : — 



^ = 13 

 and ?£ 2 =16"1 



