270 Mr Bernard , On the Unit of Classification 
have been thinking and saying shall find practical expression in 
our methods. 
I should like to see the question put to all candidates for 
Natural History appointments, “What are the aims of classifi- 
cation?” The man who is a museum official and nothing more 
would say: “In order that persons might bring their collections 
and get them named.” A more frequent answer, however, is, 
“Classification is solely for the purposes of reference.” Now the 
last answer is sometimes given by men who fully appreciate the 
vast field of research which the theory of Evolution has opened up. 
“ Continue the old fashion of naming,” say they, “but don’t call it 
classification in any high sense. The gradual discovery of genetic 
affinities by the construction of evolutionary series must be under- 
taken separately with special formulae for that kind of work.” 
This is the attitude which Dr Sharp took up in a pamphlet 
published nearly thirty years ago 1 , and, if I understand him 
correctly, it is that which Professor Lankester seemed inclined to 
endorse at one of the meetings of the Linnean Society given up 
to the discussion of my proposals. 
But, the more I think about it, the less do I find it possible to 
fall into line. Dr Sharp is of course perfectly correct in main- 
taining that we can make no satisfactory classification of forms 
which we take up, perhaps for the very first time, in order to 
describe them and give them names for reference. Their real 
positions in the evolutionary series must be left to future gene- 
rations of zoologists to find out. But at the same time it seems 
to me hopelessly unpractical, and indeed very undesirable, to 
separate the naming of animal forms for reference from at least 
some attempt at classification. We should have no means of 
knowing what forms had been named, and what not, but for the 
assistance which a classification gives us of running them down. 
This, however, does not completely overthrow the position of 
those who are inclined to advocate that naming and classification 
should be kept quite apart. For it is still possible to recommend the 
continuance of the present rough and ready method of classification 
for the purposes of mere naming for reference fully aware all the 
while that the classification in no way corresponds with what we 
now mean b}^ natural classification. Such advice can hardly be 
defended. It says : Let us start our work on lines which we 
know to be faulty, in order that we may do it later all over again 
on a better principle. 
This of course is the very last thing that Dr Sharp meant. 
The real strength of his position lies in the fact that the only 
available formula for classification implies, by the very terms 
1 Object and Method of Zoological Nomenclature. London, 1873. 
