NEw YorK AGRICULTURAL EXPERIMENT STATION. 115 
System of nomenclature— Throughout this bulletin the cultures 
have been designed according to both the Society card and the sys- 
tem of Conn. This is not the place for an extended discussion of 
the relative suitability of the reactions chosen for separating the 
groups or the types in the two systems. Any comprehensive study 
of that subject must be based upon an entirely different set of data 
from those accumulated in this work. On the other hand the 
simultaneous application of these two systems throughout this study 
has provided a good basis for comparing their relative utility in con- 
nection with such research. 
There is practically no difference in the number of subcultures 
required to determine the group number or the type name. The 
group number may be obtained from data upon four different cul- 
tures (the new card calls for six) and in determining the type name 
seven cultures may be needed although a smaller number often 
suffices. For accurate results these cultures should be made in 
triplicate. 
However, there is a marked difference in the number of detailed 
observations upon these cultures required by the two systems, those 
called for in fixing the group number being less than one-half as 
numerous as those necessary in separating the types. In recording 
the observations and in arranging the results for rapid comparison 
the present Society card leaves little to be desired. 
In applying the classification of Conn it was observed that while 
there usually were uniform results different workers would some- 
times include a culture under different types and even the same 
worker, on different days, would classify some cultures differently. 
Similar experiences have been noted by botanists when using a 
like system‘of classifying higher plants. 
With the Society card, the facts having been determined, the new 
cultures fell into their allotted places with mathematical precision. 
The classification permitted only a single interpretation and errors 
in judgment were well nigh impossible. Not only was the accuracy 
much increased but the time required both to do the classification 
and to detect the duplicates was markedly shortened. 
In the actual results of the classification by the two systems there 
is not a wide difference. ‘The flora as determined is divided into 33 
groups or 22 types. The classifications however are not strictly 
parallel. Bact. 212.22300 includes a section of Bact. lactis acidi and 
also Bact. lactis acidi D. However Conn considers Bact. lactis 
acidi D. as a variation of Bact. lactis acidi rather than as a distinct 
