66 FRED W. TANNER 



ing the characteristics of bacteria has not met with entire ap- 

 proval on the part of all bacteriologists. One of those who 

 was instrumental in introducing this numerical system has 

 said that it was a thing he regretted. Whether it is, or will be, 

 of any value may be determined after it has been used to study 

 various groups as has already been done in a few cases. 



Harding (1910) reported the study of the Pseudomonas cam- 

 pestris group. For each strain the same group number was 

 obtained which would indicate that the group number was well 

 fitted to these bacteria. This author believes that it may not 

 ''carry the separation to a group synonymous with the ordinary 

 conception of species." The statement .seems to have been 

 borne out by later investigations. 



Harding and Prucha (1908) used the Society's chart in the 

 study of the flora in cheddar cheese. In their conclusions they 

 state that this method of recording the reaction of cultures is a 

 marked advance in technique and that changes in the cheese 

 flora may be traced more accurately by its use. The sum and 

 substance of their opinion of the chart in its application to the 

 study of cheese bacteria is that it is a valuable means to an end. 



Harding, Morse and Jones (1909) in their study of soft rot 

 organisms use the group number in studying their strains. By 

 examining their data it is apparent that the group number 

 is very valuable in studying soft rot organisms, as noted by 

 Harding in 1910 when the same group was studied. 



Conn (1906) in his classification of dairy bacteria used a group 

 number, but does not draw any definite conclusions with regard 

 to its value. He still retained names for the organisms which 

 he studied. 



More recently, H. J. Conn (1915) has studied 130 cultures of 

 B. suhtilis by means of the society chart. In selecting these 

 cultures one-half of the determinations represented in the group 

 number were fulfilled because they were implied in the definition 

 of B. suhtilis. His conclusions are that different group numbers 

 do not always represent different species, and that better meth- 

 ods for making these ten determinations should be devised. 



Edson and Carpenter (1912) used the group number in study- 



