REPORT ON THE TETRACTINELLIDA. 
CV 
still have to be shown that mutual fertility in all cases served to define groups of equal 
morphological value ; in fact, this test is a sort of last defence of the notion that species 
have, generally speaking, any other existence than the convenience of the describes 
The matter may be most readily considered by supposing that the systematist had 
before him for classification all the animals that have at any time existed ; transitional 
forms between what are now accepted as the best of species would be as common as 
leaves in Val Ombrosa ; yet classification would be as possible as ever, and not quite so 
uncertain as at present. But to a great extent the classification would be quantitative, 
and the groups resulting would have more equal value than those which, with our 
fragmentary material, we can now attempt to establish. Thus, with a continuous 
spectrum, in which all the tints pass insensibly into each other, the physicist is still 
able to speak of red and blue, though had he nothing else than the mere sense of colour 
to guide him he would have to agree upon conventional lines of demarcation ; yet can 
any one assert that the results that would be so obtained would be less natural than 
such as would follow from a classification founded on a spectrum, produced by a prism 
with an absolute selective absorption for the very middle of the red, blue, and yellow 
regions '? And how do we know that the interruptions in the series of animal forms 
produced by the action of natural selection are a whit more natural than that we have 
supposed in the illustration just suggested ? According to Darwin it is the transitional 
forms which tend to become extinct under the action of natural selection ; no doubt, 
but how are we to distinguish between transitional forms, where all are transitional at 
their inception. Is it not by the action of natural selection in causing them to become 
extinct I and is this not somewhat arguing in a circle ? To me, therefore, a species is 
nothing more than a collection of more or less similar individuals, distinguished from 
other similar groups by any constant difference, however trivial, say in the details of 
form of a particular spicule, or by the relative size of the spicules in the two 
assemblages. It may be remarked that by the use of the qualification constant I revive 
that of the absence of intermediate forms under another form. But that is not really 
so, and the qualification might be omitted without affecting the definition, for if the 
character is not constant it might as well not exist. There is, however, in my mind 
a proviso that ought not to be unconsciously disregarded ; this is the geographical 
distribution as an element in the definition of a species, and this evidently derives its 
value from the underlying belief in the doctrine of descent, for if the descendants of 
two originally similar parents, one of which had its home in the Pacific and the other 
in the Atlantic, do not differ from each other in any discoverable character, we have 
nothing for it but to class them together as a single species ; should they, however, 
differ in the slightest definable particular, we shall readily make use of this to indicate 
the difference in their descent. Should now an exactly intermediate form be discovered 
in the Indian Ocean, this will not destroy the distinction between our two species, we 
(ZOOL. CHALL. EXP. PART LXIII. — 1888.) Err 0 
