MarcH 8, 1912] 
express it in numbers, you know something 
about it; but when you can not measure it, 
when you can not express it in numbers, 
your knowledge is of a meager and unsatis- 
factory kind; it may be the beginning of 
knowledge, but you have scarcely in your 
thoughts advanced to the stage of science.’’ 
Bacteriology must then become quantita- 
tive before the real foundations of the sci- 
ence are laid, before it can take its 
proper place among the sciences. And it 
can not become quantitative until we can 
measure its reactions in media of known 
chemical composition, by methods as exact 
and definite as any known to the chemist. 
We are always dealing, it is true, with liv- 
ing protoplasm, but if we place it under 
definitely determined environment, recent 
experiments lead us to believe that living 
substance always responds in a perfectly 
definite way both qualitatively and quanti- 
tatively. 
The very complexity of the society’s 
eard for the description of bacterial spe- 
cies is to me its own condemnation. It is 
our admission of ignorance. Are not the 
real diagnostic characters of a species lost 
in a maze of unessential characters? 
Aside from the morphological data, which 
I believe is all-essential, the cultural and 
biochemical data could, I am sure, be 
simplified to a very considerable extent as 
far as species differentiation is concerned. 
The forms of cultures and colonies are but 
functions of the morphology and methods 
of subdivision of the organisms themselves, 
as has been so well shown by the recent 
work of Graham-Smith2 We might then 
eliminate from the card many of the com- 
plex descriptions of cultures and colonies 
on agar and gelatin, retaining perhaps the 
agar streak and the gelatin stab, and sub- 
?Graham-Smith, G. S., ‘‘The Division and Post- 
fission Movements of Bacilli when Grown on Solid 
Media,’’ Parasitology, 3, 1910, 17. 
SCIENCE 
359 
stitute therefore information in regard to 
their determining factors. And then 
among the biochemical data if we could 
eliminate all but a few deep-seated physio- 
logical characters which are accurately 
measurable, and which can be easily de- 
termined by means of accurate chemical 
tests on synthetic media of known compo- 
sition, we would have a simple, accurate, 
all-sufficient description of a bacterial 
species. 
One of the weakest parts of bacteriology 
to-day is its taxonomy. Our methods of 
classification of bacteria are practically 
the same to-day that they were in the 
earliest days of the science. Migula, it is 
true, systematized the scheme of classifica- 
tion to a certain extent, Chester contrib- 
uted to its more accurate terminology, and 
the Winslows gave it a new impetus by 
the introduction of the methods of biom- 
etry. But when we think of the thousands 
of described species among which but 
two or three genera are recognized, it 
must be apparent at once that some of our 
generic names are seriously overworked. 
In other biological sciences the classifica- 
tion into species, genera, families, orders, 
classes, etc., is not only of great conveni- 
ence, but it also expresses for us something 
of the relationship of the different groups, 
something of their probable ancestry and 
line of evolution. I see no reason why the 
bacteria should not be classified in the 
same way. The bacteria are not excep- 
tions to the general biological laws. Varia- 
tion, selection, heredity, are the factors of 
evolution here as elsewhere. It is true 
among the unicellular forms we are free 
from many of the complications which 
enter into our discussions of the origin of 
species among multicellular animals and 
plants. Sexual reproduction is absent, 
there is no differentiation into germplasm 
and somatoplasm to prevent the acquisi- 
