69 



under water fiilh' two days; se\eral feet under. Any lar\ie of Pluuio)! 

 would have been in wel), liihernatins^, the month l)einj4 August, and no 

 doubt the last one was drowned. And from 1878 to the present time not 

 one butterfly have I seen. About ten years ago I set a plant of Chelonc 

 glabra in the garden, in order to have food at hand when needed, but 

 it had ne\er been visited by a Phaeton butterfly, and as I lixe on a hill, 

 oft' the river bottom, I did not suppose the butterfly would ever find it. 

 But on 25th June I noticed a little knot of twisted leaf fastened by a 

 web as large as a filbert, on the jjlant, and knew it for PJiaeton. How 

 a stray butterfly should find that single plant, entirely out of its range, 

 is more than I can understand. PJiaeton, unlike Cha/cedon, is restricted 

 in its food-plant to a single species, Che/one glabra, or possibly two, 

 the other being Lonicera ciliata, but on what authority this last has 

 been given I do not know. Che/one is the food-plant in this region. 

 Later, same day, the larviie were observed outside, extending the web, 

 and they were in first stage, very recently out of ag^. On 27th the 

 day began with showers, but cleared up in afternoon, and all of a sud- 

 den the whole colony was found actively at work. I entered in my 

 note-book that I believe the unusual stir was owing to a change of 

 weather for the worse which these creatures had a prevision of, for I 

 had noticed some years ago, that before a storm great efforts were made 

 by Phaeton larvae to put the webs in order. Two days later the rain 

 came, and this web was tight and all larvae housed, in contrast to the 

 web of Chalcedon on same plant, which was broken by the rain and 

 wind, and the larvae were out in the rain trying to repair it. By 7th July 

 they had passed second moult. After the first moult, which took place 

 inside the original web, the larvae moved up the stem and made them- 

 sehes a new one quite at the top, .and there passed second moult. 

 Then they extended the web. In pleasant weather they worked ac- 

 ti\ely, but when rain fell they were all inside. And whatever damage 

 was done they repaired it as soon as sunshine came. On loth July 

 the web was beaten in pieces by a storm, and next day the larvae were 

 lying about in great clusters on the leaves. But during the morning 

 they set at work and built a new web, evidently intended as the final 

 one, and for hibernation. Work went on daily, when sunshine per- 

 mitted, and by 13th the larvae were inside it. It was double, and three 

 round holes were left on the sides for egress. Early on 14th the rain 

 fell in torrents, and I expected to find the web dashed in pieces. But 

 it was not. At noon not a larva was to be seen, and the web was un- 

 injured. I looked at it daily, but saw no larva thereafter. The holes 

 spoken of were closed by an inner coating, plain to be seen. On 26th 

 I brought the web to the house and cut it open. It was double, the 

 inner part being thin, the outer of a dense texture like thin silk; in 

 parts it was triple. There were the casts of face and body made at 

 third moult, and about fifty caterpillars, H\'elv enough. I removed 



