which was frequently seen under the centre of the tests. This object, which is shown in the accompanying plate, fig. 1, I took to be the pupa of the male Ctenochiton.

Later on, in September, I found other pupæ in a more advanced stage, for I was able to detach them easily from the test of the scale. They now resembled the pupæ of Eulophus nemati, a common Hymenopterous insect. One of them is shown in my fig. 2, where it will be seen that the form of the head, with its bars or stripes, and the spurs on the end of the tibiæ, seem to refer the insect to the genera Eulophus or Encyrtus.. In Eulophus, indeed, the antennæ are branched, but this could not be distinguished in the pupa stage.

In October, when searching for more specimens of Ctenochiton in order further to examine these pupæ, I came across a tree, one of the species of Olearia, on which I found numbers of them in another condition. The leaves of the tree had evidently been pierced by a leaf-mining insect, and were covered with the blisters formed by it. Some of these were closed, others open; in each of the closed ones I found the remain of the larva of the leaf-miner and the pupa of which I was in search. It was quite evident that it could not be the pupa of a male Ctenochiton. The blisters, I may mention, were on both sides of the leaves; but the orifice by which the insect escaped was always on the under side.

I was able to procure several specimens of this parasitic insect in the imago state.

According to Westwood, there are five families of parasitic Hymenoptera; but only one has all the characters of the insect I am describing. In the Evaniidæ the antennæ are straight and the wings are veined; in the Ichneumonidæ the same; in the Chalcididæ the pupa is naked; in the Chrysididæ the abdomen is oblong-ovate. There remains only the Proctotrupidæ, and to these I relegate my insect. Of the genera of this family Diapria approaches it most, by the form of the wings. The only other genus resembling it seems to be Platygaster; but, according to Westwood, the legs in this genus are “not saltatorial,” whereas the fly before me can leap pretty actively. I may observe that Westwood states that the Coccidæ are much infested by Chalcidideous parasites, of a genus to which he gives the name “Coccophagus,” and which, he says, is intermediate between Encyrtus and Eulophus. My insect cannot be this, for the antennæ have at least 12 joints, against 8 in Coccophagus, and the three terminal joints, although soldered together, do not form a club. Again, it cannot be Encyrtus, for there is no dilation of the tarsus; nor Eulophus, for the antennæ are not branched. And the thin covering of the pupa prevents it from entering the Chrysididæ.