"I have frequently," he says, "selected one of these flies for observation, and have seen their labours extend over a period of a fortnight or twenty days; sometimes only half a cell was completed in a day, at others as much as two. I never saw more than twenty cells in one nest, seldom indeed that number, and whence the caterpillars were procured was always to me a mystery. I have seen thirty or forty brought in of a species which I knew to be very rare in the perfect state, and which I had sought for in vain, although I knew on what plant they fed. "Then again how are they disabled by the wasp, and yet not injured so as to cause their immediate death? Die they all do, at least all that I have ever tried to rear, after taking them from the nest. "The perfected fly never effects its egress from the closed aperture, through which the caterpillars were inserted, and when cells are placed end to end, as they are in many instances, the outward end of each is always selected. I cannot detect any difference in the thickness in the crust of the cell to cause this uniformity of practice. It is often as much as half an inch through, of great hardness, and as far as I can see impervious to air and light. How then does the enclosed fly always select the right end, and with what secretion is it supplied to decompose this mortar?"] _Wasps_.--Of the wasps, one formidable species (_Sphex ferruginea_ of St. Fargeau), which is common to India and most of the eastern islands, is regarded with the utmost dread by the unclad natives, who fly precipitately on finding themselves in the vicinity[1] of its nests, which are of such ample dimensions, that when suspended from a branch, they often measure upwards of six feet in length.[2] [Footnote 1: In ought to be remembered in travelling in the forests of Ceylon that sal volatile applied immediately is a specific for the sting of a wasp.] [Footnote 2: At the January (1839) meeting of the Entomological Society, Mr. Whitehouse exhibited portions of a wasps' nest from Ceylon, between seven and eight feet long and two feet in diameter, and showed that the construction of the cells was perfectly analogous to those of the hive bee, and that when connected each has a tendency to assume a circular outline. In one specimen where there were three cells united the outer part was circular, whilst the portions common to the three formed straight walls. From this Singhalese nest Mr. Whitehouse demonstrated that the wasps at the commencement of their comb proceed slowly, forming the bases of several together, whereby they assume the hexagonal shape, whereas, if constructed separately, he thought each single cell would be circular. See _Proc. Ent. Soc_. vol. iii. p. xvi.]