292 Transactions, of the Society. 



least approximately, the diameter of these flagella by means of the 

 filar micrometer, and in consequence of the comparatively very 

 considerable thickness of such distended filaments it proved feasible 

 to do so, and to obtain results closely confirming those yielded by 

 extinction. A bacillus having a long and clean flagellum was 

 selected for the purpose, a Powell ^-in. objective being employed 

 with amplifier (thus practically converting it into a 3^ of N.A. 

 1 • 27) together with a special Nelson-Powell micrometer having a 

 traversing screw-setting frame and ] O-compensating ocular. With 

 this combination the selected flagellum (one of the finest on the 

 slide) appeared broad and coarse, the spider lines being in com- 

 parison extremely fine. It was found that when the upper edge of 

 the "fixed" wire was accurately adjusted by screw to just touch 

 the lower edge of the flagellum, it required a revolution of the 

 drum through twenty -one divisions to bring the top edge of the 

 moving wire in contact with the upper edge of the flagellum, one 

 drum division being equal to y^fyoo ^ n - ^hia gives a value of 

 0*00001079 in. for the flagellum in question, which, however, 

 requires antipoint augmentation * of • 00000425 in. for the W.A. 

 0'875, thus making the true diameter 0-00001504, or 6 64F9 in., 

 as against the extinction q±\^§ m - f° r flagella of approximately 

 similar fineness. In consequence of these results I venture to 

 submit there can be no reasonable doubt regarding the practical 

 efficiency of the extinction method when dealing with smaller 

 objects so minute as to absolutely defy accurate measurement in 

 any other manner, even with the most refined appliances. 



* See this Journal, 1904, p. 271. 



