A METHOD FOR INOCULATING ANIMALS WITH PRECISE 
AMOUNTS. 
By M. J. Rosexau, Director Hygienic Laboratory, JJ. S. Public Health and Marine- 
Hospital Service. 
The need has long been felt in special lines of research work for a 
good method of inoculating animals with precise amounts. In the 
ordinaiy laboratory Avork great exactness in this regard is not an 
important matter, but in standardizing diphtheria antitoxins, and in 
the physiological tests of some drugs, and in certain lines of physio- 
logical chemistry, the greatest precision is essential. 
With the ordinaiy methods in use in most laboratories, there is an 
unavoidable loss, resulting in an error of from 1 to 8 per centum. 
The usual method consists of first measuring the dilution, or the sub- 
stance, into a graduate or other suitable vessel, and then drawing it 
up into the syringe, from which it is inoculated into the animal. 
There are tivo principal sources of error in this practice, (1) the loss 
in the glass and (2) the loss in the syringe. 
The loss in the glass . — If the fluid is drawn up into the syringe from 
a glass container a certain amount remains behind. In a series of 
18 careful weighings this loss was found to ai^erag'e about one-half of 
1 per cent, using L c.c. of fluid. 
The loss from this source was determined b}" weighing a chemically 
clean conical graduate, then adding I c.c. of distilled water, taking care 
not to wet more surface than necessary, and again weighing to 
determine the amount of water. As much of the water as possible is 
drawn up into the SAuInge, tilting the graduate so as to get the last 
drop that drains to the bottom. The graduate is again quickl}^ 
weighed before evaporation takes place to any appreciable extent. 
The difference between the weight of the wet glass and the diy glass 
of course gives the loss in the glass. The average loss in I c.c. was 
found to be 0.0192; that is, nearly two one-hundredths of a cubic cen- 
timeter, or about one-half of 1 per cent (0.18 per cent) of the amount 
of water used. Of course when less than I c.c. of ffuid is used the 
percentage of loss increases in proportion. 
3 
