1896.] The Bacterial Diseases of Plants. 637 
as answering all the requirements of systematists, but such is 
far from being the case when it comes to the description of 
bacteria. These minute organisms, which are among the low- 
est and simplest forms of living things yet discovered by man, 
are, within the commonly accepted generic limits, so morpho- 
logically similar as very often to be indistinguishable with any 
certainty even under the highest powers of the microscope. 
As supplemental, therefore, to morphology, and even in many 
cases as a complete substitute for it, we must have recourse to 
Biology, viz. to the behaviour of the living organism under a 
variety of known, artificially prepared conditions, such for ex- 
ample as the peculiarity of its growth on various culture media, 
its thermal death point, its ability to ferment various sugars, 
the chemical products of its growth, its pathogenic power, ete. 
Morphologically identical organisms often differ so widely and 
constantly in their biological peculiarities that there can be no 
question about their being distinct species, or as to the real 
value of this means of classification. Probably it also has 
value, hitherto overlooked, for the differentiation of higher 
plants and animals, especially for determining the limits of 
polymorphic or closely related species. 
It is not my intention in this place to mention all the biolog- 
ical tests which should be applied to any species for its proper 
characterization. These are being added to constantly by an 
army of trained workers in all parts of the world, and my own 
views of what is at present necessary, or at least highly desir- 
able, will be sufficiently evident in what is to follow. Very 
likely, also, as knowledge increases, some of the tests which are 
now generally held to be important will be shown to have little 
specific worth. 
This, however, appears to be a good place to insist on accur- 
acy in all the details of bacteriological work, especially in the 
preparation of culture media, and on explicitness of statement 
so that other investigators may know just what was done and 
how it was done, and thus be able to repeat the experiment. 
_ When all details of work are suppressed the inference, naturally 
enough, is that the writer was ignorant or else that he desired 
to conceal something not specially to his credit, and which if 
