was not satisfied. I put in, therefore, two more at 5 P.M. At 5.30 they were all right; at 5.45, ditto, one being almost cleaned. At 6 one was all right; the other was no longer recognizable, having been quite cleaned. At 6.30 also one was quite at home; the other could not be distinguished. At 7 both had been completely cleaned

The following day I marked another, and put her in at 6 A.M. At 6.15 she was all right among the others, and also at 6.30, 7, 7.30, 8, and 9.30, after which I could no longer distinguish her.

Again, on the following day I put in another at 6.45 A.M. At 7 she was quite at home, and also at 7.15, 7.30, 8, and to 9.30, after which I did not watch her.

To test the mode in which the ants of this nest would behave to a stranger, I then, though feeling no doubt as to the result, introduced one. The difference was very striking. The stranger was a powerful ant; still she was evidently uncomfortable, started away from every ant she met, and ran nervously about, trying to get out of the nest. She was, however, soon attacked.

Again, on October 1 some pupa of Lasius niger were placed in a glass with five ants from the same nest.

On December 8 I took three of the ants which had emerged from these pupa, and at midday put them back into their old nest, having marked them by nick

ing the claws. Of course, under these circumstances we could not watch the ants. I examined the nest, however, every half hour very carefully, and am satisfied that there was no fighting. The next morning there was no dead ant; nor was there a death in the nest for more than a fortnight.

Looked at

December 21.-Marked three more in the same manner, and put them in at 11.15 A.M. the usual intervals, but saw no fighting. The next morning there was no dead one outside the nest; but I subsequently found one of these ants outside, and nearly dead. I am, however, disposed to think that I had accidentally injured this ant.

December 23.-Painted three, and put them in at 10 A.M. At 11 they were all right, 12 ditto, 1 ditto, 2 ditto, 3 ditto, 4 ditto, 5 ditto. At 3 I put in three strangers for comparison: two of them were soon attacked; the other hid herself in a corner; but all three were eventually dragged out of the nest. I found no other dead ant outside the nest. for some days.

December 29.-Painted three more, and put them in at 10.30 a.m. At 11 they were all right, 12 ditto, 1 ditto, 2 ditto. During the afternoon they were once or twice attacked for a minute or two, but the ants seemed soon to perceive the mistake, and let them go again. The next morning I found one dead ant, but had no reason to suppose that she was one of the above three. The following morning there was again only one dead ant outside the nest; she was the third of the

strangers put in on the 23rd, as mentioned above. Up to January 23 found no other dead one.

January 3, 1879.-Painted three more, and put them in at 11.30 A.M. At 12 two were all right: we could not see the third; but no ant was being attacked. 12 ditto. 1, all three are all right; 2 ditto; 5 ditto. As already mentioned, for some days there was no dead ant brought out of the nest.

January 5.-Painted three more and put them in at 11.30 A.M. At 12 two were all right among the others; I could not find the third; but no ant was being attacked. 12.30 ditto, 1 ditto, 2 ditto, 4 ditto. On the following morning I found two of them all right among the others. There was no dead ant.

January 13.-Painted three more and put them in at 12.30. At 1 they were all right. 2 ditto. 4, two were all right; I could not see the third, but she was not being attacked. The next morning, when I looked at the nest, one was just being carried, not dragged, out. The ant carried her about 6 inches and then put her down, apparently quite unhurt. She soon returned into the nest, and seemed to be quite amicably received by the rest. Another one of the three also seemed quite at home. The third I could not see; but up to January 23 no dead one was brought out of the

nest.

January 19.-Marked the last three of these ants, and put them into the nest at 9.30 A.M. They were watched continuously up to 1. At that time two of

them had been almost completely cleaned. One was attacked for about a minute soon after 11, and another a little later; but with these exceptions they were quite amicably received, and seemed entirely at home among the other ants.

Thus every one of these thirty-two ants was amicably received.

These experiments, then, seem to prove that ants removed from a nest in the condition of рирæ, but tended by friends, if reintroduced into the parent nest, are recognised and treated as friends. Nevertheless the recognition does not seem to have been complete. In several cases the ants were certainly attacked, though only by one or two ants, not savagely, and only for a short time. It seemed as if, though recognised as friends by the great majority, some few, more ignorant or more suspicious than the rest, had doubts on the subject, which, however, in some manner still mysterious, were ere long removed. The case in which one of these marked ants was carried out of the nest may perhaps be explained by her having been supposed to be ill, in which case, if the malady is considered to be fatal, ants are generally brought out of the nest.

It now remained to test the result when the pupæ were confided to the care of ants belonging to a different nest, though, of course, the same species.

I therefore took a number of pupa out of some of my nests of Formica fusca and put them in small

glasses, with ants from another nest of the same species. Now, as already mentioned, if the recognition were effected by means of some signal or password, then, as we can hardly suppose that the larvæ or pupa would be sufficiently intelligent to appreciate, still less to remember it, the pupa which were intrusted to ants from another nest would have the password, if any, of that nest, and not of the one from which they had been taken. Hence, if the recognition were effected by some password or sign with the antennæ, they would be amicably received in the nest from which their nurses had been taken, but not in their own.

I will indicate the nests by the numbers in my note-book,

On August 26 last year I put some pupa of Formica fusca from one of my nests (No. 36) with two workers from another nest of the same species. Two emerged from the chrysalis state on the 30th; and on September 2 I put them, marked as usual, into their old nest (No. 36) at 9.30 A.M. At 9.45 they seemed quite at home, and had already been nearly cleaned. At 10.15 the same was the case, and they were scarcely distinguishable. After that I could no longer make them out; but we watched the nest closely, and J think I can undertake to say that if they had been attacked we must have seen it.

Another one of the same batch emerged on August 18, but was rather crippled in doing so. On the 21st I put her into the nest (No. 36). This ant was at once