carefully protected from the access of other insects until about the middle of June, when male and female specimens of Spathogaster baccarum were produced. The sexes united at once, and the females were then isolated and placed in captivity, each with its little oak tree. They soon laid their eggs in the leaf buds, and thus gave rise to the winter galls, which, in the following spring, produced a brood of the parthenogenetic female Neuroterus lenticularis.

He has made similar careful observations on many other species, and he gives the following table to exhibit his results:

In the following four species no males were discovered, but the parthenogenetic females gave birth to females like themselves:

Aphilothrix seminationis.

Aphilothrix quadriliniatus.

Aphilothrix marginalis.

Aphilothric albopunctata.

These are all of them insects which form galls on oak leaves, but Adler finds that the same power to lay parthenogenetic eggs exists in some other wasps. Pteromalus puparum lays its eggs in the bodies of butterfly larvæ, and thus gives birth to both males and females. The sexes are so different that there is no difficulty in separating them as soon as they are born. Adler found. that females which were thus isolated, and which were shown by microscopic examination to be virgins, nevertheless laid eggs as soon as a caterpillar was furnishedthem.

Among 206 females which hatched from these eggs there were only 9 males, so that there is, in this species, a strong tendency for parthenogenetic eggs to produce females.

In the rose-gall-wasps Adler finds that the males are very rare, about one to fifty females, and he believes that they are superfluous, since the females in two species, Rhodites rosa and Rhodites eglanteriæ are perfectly parthenogenetic, giving rise to parthenogenetic female offspring.

The instances of parthenogenesis in larval or immature insects are extremely interesting, but as they will be referred to at some length in another place I will not dwell upon them at present, as the cases which have been given are enough for our purpose, which is simply to show the satisfactory and exhaustive character of the proof that unfertilized eggs do in many animals develop and give rise to organisms which are in all respects like those born from fertilized eggs.

In Nematus ventricosus the males are not uncommon,

but Adler has verified Siebold's statement that in this species parthenogenesis of the ordinary females is not at all infrequent.

Although parthenogenesis is more frequent among the insects and crustacea than it is in other animals, it is not confined to these groups.

Cohn has given good reasons (Zeit. f. Wiss. Zool., xii., 1863, p. 197) for believing that among the Rotifera the summer eggs, which give rise to both males and females, are parthenogenetic; while the winter eggs, which hatch into females exclusively, are the only ones which are fertilized. There is no reason for doubting the correctness of this conclusion, but it has not been placed beyond the possibility of all doubt, as is the case with so many insects.

Many observers have thought that they have found evidences of parthenogenesis in groups of animals where such an occurrence would be very exceptional, but in most of these cases there is much chance for error. Thus it has been stated that the eggs of echinoderms sometimes develop without impregnation, but when we recollect that both male and female echinoderms in most cases discharge their reproductive elements into the water, we can see that it must be almost impossible to state that the sea-water in which the eggs are placed contains no spermatozoa of the same species. Dr. J. M. Wilson has recently undertaken some experiments on this point at my suggestion. He fertilized a lot of eggs from one of our common sea-urchins, Strongylocentrotus, with male fluid from another of a distinct genus, Arbacia. A lot of Arbacia eggs were fertilized with a male Strongylocentrotus, a lot from each form with fluid from a male of the same species, and eggs from each species were placed in water without fertilization.

In all six cases the eggs gave rise to normal embryos; but that this was really due to the presence of spermatozoa in the water, was shown by the fact that no such surprising result followed in a second set of experiments where especial effort was made to get pure sea-water. Many of the recorded cases are open to the same objection; and in other cases, as in the virgin sou referred to by Bischoff, there seems to be some doubt whether the ova were really undergoing development; but Oelacher's observations on the eggs of a virgin hen (" Die Veränderungen des unbefruchteten Keimes des Hühnereies im Eileiter und bei Bebrütungsversuchen," Zeit. f. Wiss. Zool., xxii., 1872, p. 220) seem to show that the hen's egg does have the power to pass through the first stages of development whether it is impregnated or not.

The instances of parthenogenesis which I have given show that this power may be independently acquired by animals which cannot possibly inherit it from a common source. In the vast majority of insects, and in the majority of the crustacea, the egg does not show the slightest tendency to develop before it is fertilized. It is true that in the case of the crustacea the evidence for this statement is almost entirely of a negative character, for no one has ever shown by experiment on any considerable number of species that the female cannot lay fertile eggs when the access of a male is prevented, but in many insects we know from actual observation that the eggs die soon after they are laid, unless they are fertilized; and we know enough of the breeding habits of crustacea to feel confident that parthenogenesis is exceptional among them, just asit is among insects.

We must, therefore, conclude that if we could retrace the course of evolution of any parthenogenetic animal we should be led back to an ancestral form which never

manifested any such power. It is impossible to believe that Daphnia and the honey-bee have inherited from a common parthenogenetic ancestor the power to produce fertile unimpregnated eggs, for the one form is much more closely related to normal insects and the other to normal crustacea than they are to each other. We may therefore state with confidence that the power has been independently acquired by many animals.

In the second place, we must admit that parthenogenetic ova are true ova in every sense: they are developed in an ovary like other eggs, and in many cases, as in those butterflies which are occasionally parthenogenetic, the very eggs which usually require impregnation may in rare instances develop without it. Weismann has made very careful examination as to the origin of both kinds of eggs in Leptodora, a water-flea related to Daphnia (Ueber die Bildung von Wintereier bei Leptodora hyalina,” Zeit. f. Wiss. Zool., xxvii., 1876), and he finds that while there is some difference in the mode of origin of the winter eggs, which do not develop unless they are fertilized, and the summer eggs, which are parthenogenetic, the difference simply consists in the amount of nourishment which they receive in the ovary. In each case certain ova degenerate and are used up by the others as food, and a winter egg thus absorbs a greater number of these embryonic ova than a summer egg does; but Weismann's observations show that each of them is in all respects a true ovum, and that they are perfectly homologous with each other.

In some cases, as in some of the wasps described by. Bassett and Adler, the animal which is born from a parthenogenetic egg differs considerably in structure from that which is born from a fertilized egg; but in other cases, as in butterflies and moths, there is no such