NISONIADES TAGES. 273 



prefer to get the material to form a complete enclosure, so that only 

 the edges have to be held together, but it must often happen, naturally, 

 that they fail in this, as they are perfectly prepared to close in the 

 deficiency with a web of silk ; one or two escaped larvae had used this 

 expedient, and two that were in pill-boxes spun in an angle, and the 

 whole enclosure from bottom to side of pill-box was made of silk. This 

 is of special pattern, being a network of which the bits are very dense, 

 but it is full of more or less circular holes of various sizes. Unless 

 d isturbed in some way, the larva pupates in spring in its hybernating 

 cocoon. 



Laeva. — First instar (newly-hatched) : Rich red-brown in colour, 

 with black head and prothoracic plate. [Comparing it with Scudder's 

 figures (Butt. New Engl., pi. lxxiii), it agrees in form with fig. 2 

 (Thanaos lucilius), and in the hairs, etc., of the prothorax, but the 

 plate is black. The head agrees very closely with fig. 8 (Epargyreus 

 tityrus) as to sculpturing and hairs ; the hairs of the meso- and meta- 

 thoraxand of the abdominal segments also agree very closely with fig. 

 8, if we suppose tubercle iii to be omitted in fig. 8, on a dorsal view 

 like this, it is to be detected with difficulty. In tages also, on the 9th 

 abdominal segment, the posterior dorsal hair is a short club, like the 

 other, not a bristle, and iii is present, but not iv; on the 10th abdominal 

 are two clubs and two bristles on either side.] l*7mm. long. Head 

 black, minutely pitted, with fine minute clubbed solitary hairs at usual 

 sites of tubercles ; 8 on either side, regularly disposed, being easily 

 visible on dorsal view. The prothorax is divided into several subsegment s ; 

 the 1st has two dorsal (equally spaced) hairs on either side, the 2nd 

 subsegment with no hair behind middle one, the 3rd with a hair behind 

 outer one of 1st; behind these subsegments is a subsegment without any 

 hairs ; the prothoracic plate has four hairs on each side, two in line with 

 i, and two with ii ; the plate is divided into three transverse pieces, 

 and hairs are, inner on the 1st and 2nd, and outer on the 1st and 3rd ; 

 lower down is a bristle. The meso- and metathorax have each two 

 (equally spaced) hairs on either side of dorsum of the 2nd subsegment, 

 also one marginal hair; one subsegment being in front of this and two 

 behind, without hairs. The abdominal segments are divided into 5 

 subsegments, of which the first is the largest, and unites with the 

 second at about the level of iii ; the rest go down to the lateral flange ; 

 the 1st subsegment carries tubercle i, the 3rd carries ii, whilst iii and 

 the spiracle below it are in about the line of division between the 1st 

 and 2nd subsegments, which (as just noted) fails at this level ; below 

 the spiracle is a faint flange, then a more marked one carrying iv and 

 v, the latter fractionally the higher, and close to front of segment ; the 

 9fch abdominal segment shows two subsegments, the 1st carrying the 

 setae of i (?), the 2nd, ii and iii (?), and a lower one v (?) where the 

 subsegmentation ceases ; the 10th abdominal carries one clubbed hair 

 dorsally, one laterally, with three marginal (ordinary) hairs on each 

 side, behind, on the flange. Seen from above, the setae arising from 

 iv and v form a fairly regular marginal row round the larva. The 

 hairs arising from all these abdominal tubercles are really short glassy 

 clubs (see pi. vi., fig. 1) ; the little clubs seem wrinkled at their thick 

 ends, they are only a few times longer than thick, except those on 

 iv and v which are longer, perhaps 0*02mm. long, and are more 

 definitely knobbed hairs. They average about three times as long as 



