MKLANAEGIA LACHESIS. 303 



seem at all anxious to feed. I think it likely that they do not at once 

 take food, but may wait for some weeks until the grasses have been 

 freshened by rain. If they do eat during this period it is only very 

 little. These did not touch any of the grasses I was able to give them 

 during the first ten days of their lives, but I think they nibbled some 

 Featuca later on. On September 20th, I put them on a tuft of this 

 grass growing in a pot, and brought them from England to Hyeres. 

 Having seen no trace of them or of their feeding for some weeks, I 

 began to think that all had perished, but about the third week in 

 October they must have become active, for, on October 29th, I exammed 

 the plant and found eight larvae up on the leaves. They were still in 

 the first stadium and quite small, but a tinge of green on the fore 

 part of the body proved that they had taken food, and freshly-eaten 

 leaf-ends were to be seen. After that date they w^ere several times 

 observed up on the grass in the daytime. When startled, they assume 

 an attitude unusual in Satyrid larvse. The slightest jar or shock will 

 cause them to loosen their hold of the grass stem or leaf, with all but 

 the anal (and sometimes 4th ventral) pair of claspers, curve the body, 

 and bring the head down underneath until it is below the 2nd abdominal 

 segment (see pi. xiv., fig. 3). The curious position assumed by the young- 

 larva of Satynis var. allionia when startled is quite a different thing. 

 Examined on November 8th (a fine day after a week of rainy weather), 

 they were found healthy and making slow progress. I noted them of 

 a straw-grey colour, with brownish lines, and generally a greenish 

 tinge about the thoracic segments. 



Description of the ovum (made from three empty shells, two of 

 which have been partly eaten by the larva). — Shape, roundish, base 

 broad, Avith a central depression, top broadly rounded ; height and 

 breadth nearly equal, l-05mm.-lmnQ. Appearance, chall^y, with a very 

 slight tinge of bufl:; eggshell rather thick and opaque. There are 

 about 33-34 main ribs, which become distinct about -i way up the Qgg, 

 and terminate at the edge of the broad, low, dome-like top. The 33 

 or 34 main ribs I counted reach this point, but there are some short, 

 intermediate ones reaching only to the middle of the egg. The 

 depression of the base is slightly roughened. I cannot make out 

 any cell system in it. All around, and extending up to the lower limit 

 of the ribbing, are numerous elevations or points, giving this part of 

 the QOQ a thickly pimpled look. Between these points run low, rather 

 broad and ill-defined, walls, enclosing shallow cells. The ribs generally 

 merge into the pimples. Something similar occurs towards the top of 

 the Qgg, when the ribs again break up into pimples and a rough cell- 

 system. The size of the cells diminish as the micropylar area is 

 approached, whilst the pimples disappear. In the centre of all is a 

 rosette of very small cells. The main ribs do not run in particularly 

 straight lines, they are rather wavy, especially towards their lower 

 extremities, where they in some cases divide into two before definitely 

 breaking up into pimples. The ribs do not stand out high from the 

 surface, as compared with those of many Satyrid ova. I should say 

 they are more pronounced than in Melananjia var. jn-odda, but are 

 much less so than in M. sylUus. The cross ribbing is very minute, 

 and the number of cross ribs great. Mr. Tonga's photo of the shells 

 is a very good one. 



Description of larva. — Fir&t stadium (made on November Sth, 



