I>rVESTIGATION OF AMERICAN STAINS 131 



both of them using different technics; although one of these two 

 (Castleman) made his tests on two different occasions, with a 

 different technic each time. 



Results were reported in two different ways. Some grouped 

 the samples in three or four categories which it has proved 

 possible to denote by four terms: excellent (E), good (G), fair 

 (F), and unsatisfactory (U). Others listed the samples in the 

 order of their excellence, in which case instead of placing them 

 in these four categories, they have been numbered 1, 2, 3, etc., 

 the low numbers indicating the best samples. These two sets of 

 symbols are used in Table 1, where the results of the different 

 investigators with the different samples are listed. 



It will be seen that there is some variation in the findings of the 

 different investigators, although not as much as in the case of 

 the other two dyes, reported below. When fuchsin was used as 

 a stain it proved difficult to pick out any one sample or any two 

 or three samples that were superior to the others. To make 

 the comparison more definite, each of the four classes was given 

 a numerical symbol, these numbers averaged, and the average 

 converted back into the class symbol again, using the grades: 

 E, E— , G+, G, G — , etc. By glancing at this average grade, 

 it will be seen that the sample from the Providence Chemical 

 Company ranks G + , none of the others ranking better than G. 

 All but three of the others rank G. 



The samples are rearranged in table 2 in the order of excellence, 

 according to their average grade, this same table listing the 

 number of times each sample was found excellent and the number 

 of times found unsatisfactory. Only three samples were re- 

 ported unsatisfactory by any one, namely those from dealers 

 A, B, and C, each being so reported by two different investi- 

 gators. These three were the only samples to have an average 

 grade lower than G, and yet two of them were reported as good 

 or excellent by some investigator. There seems no reason to 

 feel that any of the samples, with the exception of these three, 

 could not be successfully substituted for the Griibler product. 

 The superiority of the Providence sample is so slight as to be of 

 little significance. 



JOURNAl. OF BACTEHIOLOOT, VOL. VII, NO. 1 



