1

that in some cases, at any rate, the eggs do produce Dewitz even maintains that the workers young. habitually lay eggs, and explains the difference which on this view exists between the workers of ants and those of bees, on the ground hat (as he supposes) the majority of ants die in the autumn, so that the eggs laid by the queens alone would not be sufficient to stock the nest in the spring; while among bees the majority survive the winter, and consequently the eggs laid by the queen are sufficient to maintain the numbers of the community. In reply to this argument, it may be observed that among wasps the workers all perish in the autumn, while, on the contrary, among ants I have proved that, at least as regards many species, this is not the case. Moreover, although eggs are frequently laid by workers, this is not so often the case as Dewitz appears to suppose. Forel appears to have only observed it in one or two cases. nests the instances were more numerous; and, indeed, I should say that in most nests there were a few fertile workers.

In my

Among bees and wasps also the workers are occasionally fertile; but, so far as our observations go, it is a curious fact that their eggs never produce females, either queens or workers, but always males. The four or five specimens bred by Forel from the eggs of workers were, moreover, all males.

It became therefore an interesting question whether

1 Zeit. ¡. wiss. Zool., vol. xxviii. p. 586.

the same is the case among ants; and my nests have supplied me with some facts bearing on the question. Most of my nests contained queens; and in these it would be impossible, or at least very difficult, to distinguish and follow the comparatively few eggs laid by the workers. Some of my nests, however, contained no queen; and in them therefore all the eggs must have been laid by workers.

One of these was a nest of Formica cinerea, which I brought back from Castellamare in November 1875. At that time it contained no eggs or larvæ. In 1876 a few eggs were laid, of which fifteen came to maturity, and were, I believe, all males. In 1877 there were fourteen pupæ, of which twelve came to maturity, and were all males.

Again, in a nest of Lasius niger, kept in captivity since July 1875, there were in 1876 about 100 young; and these were, as far as I could ascertain, all males. At any rate there were about 100 males, and I could not find a single young female. In 1877 there were again some pupæ; but owing to an accident none of them came to maturity. In 1878 fifteen came to maturity; and fourteen were males. The other I could not find after it left the pupa skin; but I have no doubt, from the appearance of the pupa, that it was also a male.

Another nest of Lasius niger, taken in November 1875, brought in 1878 only one young ant to maturity; and this was a male.

Again, in a nest of Formica fusca, taken in 1875, though in 1876 and 1877 eggs were laid and a few arrived at the pupa-state, none came to maturity. They were all, however, either males or queens, and, 1 have little doubt, were males. In 1878 one came to maturity, and it was a male.

A nest of F. fusca, captured in 1876, did not bring up any young in 1877. In 1878 three larvæ came to maturity; and they all proved to be males. Another nest of F. fusca, captured in 1877, in 1878 brought only one young one to maturity. This was a male.

In the following year, I again carefully watched my nests, to see what further light they would throw on the subject.

In six of those which contained no queen, eggs were produced, which of course must necessarily have been laid by workers.

The first of these, the nest of Lasius niger, which I have watched since July 1875, and which, therefore, is interesting from the great age of the workers, about ten larvæ were hatched, but only four reached the pupa state. Of these one disappeared; the other three I secured, and on examination they all proved to be males. The nest of Lasius niger, which has been under observation since November 1875, produced about ten pupa. Of these I examined seven, all of which I found to be males. The others escaped me. I believe that, having died, they were brought out and thrown away.

The nest of Formica cinerea, captured at the same time, produced four larvæ, all of which perished before arriving at the pupa stage. The larvæ of males and of queens are much larger than those of workers, and these larvæ were too big to have been those of workers.

In a nest of Formica fusca, which I have had under observation since August 1876, three pupae were produced. They were all males. Another nest of Formica fusca produced a single young one, which also was a male.

Lastly, my nest of Polyergus rufescens, which M. Forel was so good as to send me in the spring of 1876, in 1879 produced twelve pupa. Eleven of these turned out to be males. The other one I lost; and I have little doubt that it was brought out and thrown away. It was certainly not a worker. As regards the first three of these рирæ, I omitted to record at the time whether they belonged to the Polyergus or to the slaves, though I have little doubt that they belonged to the former species. The last eight, at any rate, were males of Polyergus.

Indeed, in all of my queenless nests, males have been produced; and in not a single queenless nest has a worker laid eggs which have produced a female, either a queen or a worker. Perhaps I ought to add that workers are abundantly produced in those of my nests which possess a queen.

While great numbers of workers and males have

come to maturity in my nests, with one exception not a single queen has been produced.

This was in a nest of Formica fusca, in which five queens came to maturity. The nest (which, I need hardly say, possessed a queen) had been under observation since April 1879, and the eggs therefore must have been laid in captivity. The nest had been richly supplied with animal food, which may possibly account for the fact.

It is known that bees, by difference of food, &c., possess the power of obtaining at will from the same eggs either queens or ordinary workers. Mr. Dewitz,1 however, is of opinion that among ants, on the contrary, the queens and workers are produced from different kinds of eggs. He remarks that it is very difficult to understand how the instinct, if it is to be called instinct, which would enable the working ants to make this difference can have arisen. This is no doubt true; but it seems to me quite as difficult to understand how the queens, which must have originally laid only queen eggs and male eggs, can have come to produce another class. Moreover, however great the difficulty may be to understand how the ants can have learnt to produce queens and workers from one kind of egg, the same difficulty exists almost to the same extent in bees, which, as Mr. Dewitz admits, do possess the power. Moreover, it seems to me very unlikely hat the result is produced in one way in the case of

1 Zeit. für wiss. Zool. 1878, p. 101