FOR MEASURING TENSILE STRENGTHS. S91 



sion of the floating cylinder and the containing vessel, but 

 it will always be in the direct ratio of the sectional areas 

 of the floating cylinder and the ring of water surround- 

 ing it. 



The apparatus I have actually used in my experiments 

 upon textile fibres consists of an outer vessel provided 

 with a tap, made of common tin plate, standing 13 inches 

 high and about 3f inches in diameter, possessing a sec- 

 tional area of about 9*8 inches. I have used three float- 

 ing cylinders : the larger one, made of tin plate, has a cir- 

 cumference of 8 inches, and a sectional area of about 

 5*09 inches; the medium cylinder is of glass, having a cir- 

 cumference of 2*4 inches, a sectional area of 0*458 inch ; 

 and the smaller one is a fine-stemmed glass hydrometer, 

 having a circumference of 0*5 inch, and a sectional area of 

 •001989 inch. 



The ratios of the areas of these floating cylinders to the 

 difiference between them and that of the fixed cylinder 

 (being the area of the ring of water) are as follows : — 



For the larger ,. i 



For the medium i 



For the smaller i 



0-925 

 io-6i 

 4926 



Therefore, supposing these measurements correct, and the 

 fixed and floating cylinders perfectly cylindrical and even 

 throughout, i grain weight would be put upon the fibre 

 by the withdrawal of 0925 grain of water for the larger, 

 20' 6 1 grains for the medium, and 492*6 grains for the 

 smaller floating cylinder. 



But, actually, neither the fixed nor floating cylinders 

 in my apparatus are so accurately made as to justify a 

 reliance upon arithmetical results; and I have graduated 

 them by actual testing with a delicate chemical balance, 

 ascertaining the weight required to keep the floating cy- 

 linders at one constant level for various quantities of water 



SER. III. VOL. II. 2 D 



