140 GENERAL BACTERIOLOGICAL DIAGNOSIS 



bacterium. The apparent size often shows slight variations dependent 

 on the stain used and the growth conditions of the culture. Accurate 

 measurements of bacteria can only be made by preparing microphoto- 

 graphs of a definite magnification, and measuring the sixes on the 

 negatives. From these the actual sizes can easily be calculated. A 

 rough method of estimating the size of an organism is to mix a little 

 with a drop of the observer's blood and make a blood film. As the size 

 of a normal red blood corpuscle is about 7 '5 IJL, an idea of the size of a 

 bacterium can be obtained by comparing it with this as a standard. In 

 describing bacterial cultures it must be borne in mind that the appearances 

 often vary with the age. It is suggested that in the case of cultures 

 grown at from 36 to 37 C. the appearances between twenty-four and 

 forty-eight hours should be made the basis of description, and in the 

 case of cultures grown between 18 and 22 C. the appearances between 

 forty-eight and seventy-two hours should be employed. The culture 

 fluids used must be made up and neutralised by the precise methods 

 already described. The investigator must give every detail of the 

 methods he has employed, in order that his observations may be capable 

 of repetition. 



In the case of a number of pathogenic organisms, identification 

 is a comparatively easy matter. In some cases, however, great 

 difficulties arise in consequence of the existence of groups of 

 organisms presenting closely allied characters, and the difficulty 

 and importance of identification is enhanced by the fact that 

 the same group may include both harmful and innocent members. 

 Examples of this occurrence are found in the pyogenic cocci and 

 their allies, in the coli-typhoid group of bacilli, and in the group 

 of cholera vibrios. In such cases it is usually necessary to take 

 into account all the morphological and cultural reactions of an 

 organism before it can be adequately classified. Within recent 

 years attempts have been made to apply the statistical method 

 to the solution of the difficulties of the situation, and here the 

 results appear to be promising. The method has been applied 

 to the coccaceae by Winslow and Rogers, who have investigated 

 500 strains of cocci isolated from the tissues in disease, from 

 the outer surfaces of the normal human body, from water, earth, 

 and air. A great variety of properties was studied, and while 

 in each test applied wide variation was exhibited in such bacteria, 

 there usually emerged a type property to which individual 

 strains tended to approach. Thus, while the size varies from 

 "1 to 2'0 fji, out of about 350 strains examined about 115 

 measured '3 ^ and the remaining strains tended to be a little 

 below or a little above this figure. When similar lines of 

 inquiry were pursued with regard to other characteristics of 

 the organisms, it was found that important correlations could 

 be noted. Thus capacity for staining by Gram's method was 

 found especially amongst the staphylococci and streptococci as 



