_Cypris vidua_ and _Cypris reptans_ are such species. Now, although the transformation of these formerly bisexual species into purely unisexual female species appears to be complete[197], yet the females still possess a large, pear-shaped _receptaculum seminis_, with its long spirally twisted duct, which is surrounded by a thick glandular layer. This is the more remarkable as the apparatus is very complicated in the Ostracodes, and retrogressive changes could be therefore easily detected. Furthermore among insects, in the genus _Chermes_ the _receptaculum seminis_ of the females has also remained unreduced, although the males appear to be entirely wanting, or at least have never been found, in spite of the united efforts of several acute observers[198]. The case is quite different in species which retain both sexual and parthenogenetic reproduction. Thus, the summer females of the _Aphidae_ have lost the _receptaculum seminis_; and in these insects sexual reproduction has not ceased, but alternates regularly with parthenogenetic reproduction. Certainly this proof of the truth of my theory as to the significance of sexual reproduction is far from settling the question: it only renders the theory highly probable. At present it is impossible to do more than this, because we do not yet possess a sufficient number of facts, for many of them could not have been sought for until after the theory had been suggested. We are here concerned with complicated phenomena, into which we cannot acquire an immediate insight, but can only attain it gradually. But, nevertheless, I hope to have shown that the theory of natural selection is by no means incompatible with the theory of ‘the continuity of the germ-plasm;’ and, further, that if we accept this latter theory, sexual reproduction appears in an entirely new light: it has received a meaning, and has to a certain extent become intelligible. The time in which men believed that science could be advanced by the mere collection of facts has long passed away: we know that it is not necessary to accumulate a vast number of miscellaneous facts, or to make as it were a catalogue of them; but we know that it is necessary to establish facts which, when grouped together in the light of a theory, will enable us to acquire a certain degree of insight into some natural phenomenon. In order to direct our attention to those new facts which are of immediate importance, it is absolutely necessary to seek the aid of some general theory for the arrangement and grouping of those which we already possess. This has been my object in the present paper.