174 Transactions of the Royal Microscojpical Society, 
answer best, is the ‘‘ enamelled cards ” which are used by the 
printers for visiting cards. I also use a H H H H, Windsor and 
Newton or Faber pencil (with the wood cut considerably away in 
the former, the latter being obtainable in solid cylinders, which are 
slid into holders), which is brought to its final point by gentle 
rubbing on the surface of the finest ground glass, or, better still, a 
very fine hone. With these appliances a drawing was made, first, of 
the lower half of the body of the B. termo, and then, which was the 
really critical matter, a pencil mark was made over a half or two- 
thirds of the flagellum, not over the whole; for in this way the 
flagellum image and the pencil mark could be carefully compared, 
as shown at Fig. 11, where the dotted part a represents the image 
of the flagellum as seen beyond the pencil line c ; and a very close 
approximation may thus be made between them. 
Having determined that the pencil mark, as at c. Fig. 11, 
accurately corresponded to the image of the flagellum, this drawing 
was taken and magnified from five to ten diameters, the amount of 
magnification being accurately determined beforehand. By this 
means it was easy to determine the ratio existing between the 
now measurable diameter of the pencil line representing the 
flagellum and the diameter of the body. But this latter was a 
known quantity ; and therefore it was easy to determine the actual 
diameter of the flagellum. Thus at h, Fig. 11, there is an outline 
camera drawing of the lower half of a B. termo magnified 2000 
diameters, c is the pencil line corresponding to a part of the 
flagellum, the dotted line a representing the remainder, with which 
the pencil line could be compared. At Fig. 12 we have the same 
drawing magnified five diameters ; and in this condition it is quite 
easy by means of the “ screw micrometer ” to find the' ratio existing 
between the maguified image of the drawing of the body and that 
of the flagellum. In other words, it is soon seen how many 
flagellum spaces are needed to cover the diameter of the body. In 
the case before us, the bacterium drawn at Fig. 1 1 had a diameter 
of 2 (XTyth of an inch. The ratio of the flagellum to the body, as 
seen in Fig. 12, is as 10 to 1 ; and -4- 10 = 
inch, the actual size of the flagellum. 
Now I made fifty separate drawings and measurements with 
each of the four lenses ; the same conditions being observed in each 
case. The results expressed in decimal fractions are as follow, 
viz. : — 
(1) The mean value of fifty measurements made with the 
-fi^th inch objective, gives for the diameter of the flagellum 
0-00000489208. 
(2) The mean value of fifty measurements made with the 
^Vth inch objective gives 0-00000488673. 
