VETERINARY MEDICINE. 83 



tococcus, and colon by three pairs of observers, is remarkably similar, being 

 on the average less than 5 per cent with a minimum of 0.5 per cent, and a 

 maximum of 21 per cent for the chamber method, and more than 15 per cent 

 with a minimum of 1.1 per cent and a maximum of 54 per cent for Wright's. 

 The fact that less time was spent on counting ()4 squares by the chamber 

 method than 75 or 50 fields by Wright's still further demonstrates the superi- 

 ority of the former. 



" The 0.1 mm. chamber is a little less accurate than 0.02, partly owing to the 

 difficulty of correctly ascertaining to which squares the bacteria adhering to 

 the under surface of the cover slip belong and partly owing to the greater 

 amount of Browuian movement. Wright's method usually greatly underesti- 

 mates the strength of the bacterial emulsion, sometimes by 100 or even 200 

 per cent. The main reasons are, first, because large numbers of bacteria are 

 left behind at the beginning of the smear; and, second, because small groups 

 of bacteria mix up with blood platelets and leucocytes and are drawn to the 

 end of the smear, where they may be unnoticed unless specially looked for 

 with a low power. Allen's modification of Wright's method has the same 

 defects as the original ; it underestimates the emulsion to the same extent, and 

 the error is, if anything, larger. The plate-culture method is cumbersome, and 

 time is wasted in waiting for the colonies to develop. It also seriously under- 

 estimates the strength of the bacterial vaccines, first, because of the impossi- 

 bility of obtaining a homogeneous emulsion, especially in the case of the cocci ; 

 second, because even in comparatively young cultures a large number of 

 bacteria are moribund or dead and fail to develop into colonies. Working 

 with most carefully prepared emulsions, we found that coli were underestimated 

 about 5 per cent in cultures four hours old, and about 60 per cent in cultures 

 24 hours old, while the staphylococci were underestimated by 50 and 100 per 

 cent, respectively, the discrepancy being greater owing to tJie large number 

 of diplococci in the younger cultures. 



" This paper has been written because we believe it is the duty of the 

 bacteriologist to standardize and dispense his vaccines with as much care 

 as is exercised by the pharmacist in standardizing and dispensiiig his drugs. 

 Scientific A^accine therapy will be advancetl by scientific exactness." 



Observations on hemolysin production by the streptococci, II. W. Lyall 

 (Jour. Med. Research, 30 (19U), No. 3, pp. 515-532, fig. 1).—At the outset 

 stress is laid upon the fact that in spite of extensive studies on this topic our 

 understanding of the factors responsible for the grave toxic conditions arising 

 from streptococcic infections is still far from satisfactory. This work was 

 therefore conducted with a view of discovering any possible correlation be- 

 tween hemolysin production and pathogenicity, and determinations were made 

 of the hemolytic titer of several strains of streptococci. 



It was found that " the hemolytic titer does not afford any absolute criterion 

 of virulence. The production of a potent hemolysin in broth cultures is de- 

 pendent on the nature and proportions of enriching substances, the reaction of 

 the medium, and the time of incubation. The hemolysin appears to be closely 

 associated with the bacterial bodies and not to be in solution. The hemolysin 

 does not appear to bo in the nature of an enzym. The hemolysin is destroyed 

 by a temperature of 56° O. for 30 minutes. It disappears within 49 hours at 

 incubator temperature. 



" Normal sera of sheep, guinea pig, rabbit, cow, and man contain appreciable 

 amounts of antistreptolysin. Salvarsanized sera possess a markedly increased 

 inhibitory jiower on the hemolytic action of streptococci. Hemolysin produc- 

 tion is inhibited by the addition of sugar. This occurs most uniformly in the 



