Romanes, G. J. (1897). Characters as adaptive and specific (Concluded [Part 4]). In G. J. Romanes, Darwin, and after Darwin: An exposition of the Darwinian theory and a discussion of post-Darwinian questions: Post-Darwinian questions of heredity and utility (pp. 251-287). Chicago, IL, US: Open Court Publishing Co.
http://dx.doi.org/10.1037/12871-010
Our subject is not yet exhausted. For it remains to observe the consequences which arise from the dogma of utility as the only raison d'être of species, or of specific characters, when this dogma is applied in practice by its own promoters. In conclusion, it only remains to reiterate that in thus combating what appears to me plainly erroneous deductions from the theory of natural selection, I am in no wise combating that theory itself. On the contrary, I hope that I am rendering it no unimportant service by endeavouring to relieve it of a parasitic growth--an accretion of false logic. Regarding as I do the theory of natural selection as, primarily, a theory of the origin (or cumulative development) of adaptations, I see in merely nonadaptive characters--be they "specific" or other--a comparatively insignificant class of phenomena, which may be due to a great variety of incidental causes, without any further reference to the master-principle of natural selection than that in the presence of this principle none of these non-adaptive characters can be actively deleterious. But that there may be "any number of indifferent characters" it is no part of the theory of natural selection to deny; and all attempts to foist upon it a priori "deductions" opposed alike to the facts of nature and to the logic of the case, can only act to the detriment of the great generalization which was expressly guarded from such fallacies by the ever-careful judgement of Darwin. (PsycINFO Database Record (c) 2017 APA, all rights reserved)