( 442 ) 



terms "species'' and " v<aiiety," a distinction which indeed everybody silently accepts 

 who considers the enormous numbers of difierent forms of animals and plants to be 

 the outcome of divergent development, and expresses this development by the con- 

 ventional figurative tree. If the specimens represented by any given portion of a 

 branch of this tree were so constructed that under favourable circumstances they 

 would be identical with the specimens represented by a portion of another branch, 

 i.e. if any two branches could, and can, long after the common origin, merge together, 

 then branch off again, merge together again, and so on, it would be preposterous to 

 assume that this should never have happened. J5ut if we thus should have to concede 

 the possibility that the lines of ancestors of any two forms of jilants or animals, say 

 of the lion and the giraffe, were such that they first became widely divergent, then 

 identical, then again widely divergent, not only the figurative tree, but also the 

 kind of evolution it is meaiit to illusti'ate, would be pure nonsense. Therefore we take 

 it that we are actually agreed upon that part of our definition which says that from 

 a certain point a branch of the tree cannot merge together again with any other 

 branch ; now, if we call every form which has reached this degree of development 

 specifically distinct, we have an absolute distinction between species and the lower 

 degrees of development. 



The question of specific distinctness or non-distinctness is therefore twofold : 

 first one of morphological, and second one of physiological difference. As the 

 systematist is practically not able to test hy experiment the presence of the second 

 distinction, it is obvious that he never can pi'ove with certainty from the sj^iecimens 

 alone whether the distinguishing morphological characters they exhibit are of specific 

 value or not. However, we are able to arrive at a probably correct conclusion without 

 testing in each case the specific distinctness, if we take into account the way in which 

 divarication of species comes about, and if we further com[iare the characters of such 

 forms as have been tested to be specifically distinct. 



Yoi- owe present purposes it is quite irrelevant whether the causes of the trans- 

 mutation and divarication of species are those factors which are maintained by the 

 Neo-Darwinians to be the sole agents, or those which the \eo-Lamarckians consider 

 to be alone effective. Hence we shall abstain from any discussion of the much 

 contested final causes of divergent development, and shall simply ask, which is the 

 ■way that leads from variability of a species to divarication of this species into more 

 species ? Our purpose allows us to simplify the question still more and to restrict the 

 discussion to the two jioints: first, can a species develop into two (or more) species 

 without i.solation ? secondly, can isolation as such transmute one species into two 

 (or more) ? 



The most extreme kind of variability of a sjiecies logically possible is that in 

 which the varieties composing the .species are not only morphologicallv but also 

 physiologically different. As upon the occurrence of such variation Romanes's theory 

 of Phj'siological Selection, which we have had to allude to several times, is founded, 

 we may be allowed to annex our notes to a short discussion of this theory as far as 

 it bears upon our particular question. 



In order to explain the infertility or restricted fertility between different species 

 Romanes assumes that the divarication of one species into more species has something 

 to do with the occurrence of such a variation that some individuals of a given species 

 are not fertile with the rest of the species, but are fully fertile inter se. That 

 variety, though living in the same district as the normal form of the species, will 

 develop divergently, according to the theory, and give rise to a new species, as it 



