212 6RITISH BUTTERFLIES. 



8th abdominal segment, and that the larvae were attended by ants, the 

 larvae keeping out the little tubes until an ant touched them with its 

 antennae, when they were drawn in, but thrown out again as soon as 

 the ant passed on ; the larvae, however, progressed quite satisfactorily 

 when deprived of the company of the ants, the latter, therefore, not 

 appearing to be necessary to the former's welfare. It is remarkable 

 that Powell observed the "caruncles," or " evaginable processes," or 

 " fans," on the 8th abdominal segment, but missed the essential fact 

 that these were merely attractive, or guide-post, arrangements, and that, 

 whilst these were withdrawn , the ant was most probably busy at the honey- 

 gland on the 7th abdominal. It would appear that the larvae have to get 

 rid of the "honey" or "fluid" when not attended by ants, for Chapman 

 notes (i?i litt.) that, on April 15th, 1908, whilst he w 7 as looking at 

 some larvae in confinement, one exuded a clear drop of fluid from the 

 honey-gland, and that, at the time, the bases of the " fans" on the 

 8th abdominal, appeared as quite large, pale circles, with darker 

 margins (? of lenticles). Gillmer states (in litt.) that Viehmeyer 

 reports finding larvae of P. argus with Lasius alienus. It is also 

 doubtfully recorded by Torka with Formica cinerea. 



Ontogeny of larva. — First instar (April 1st, 1893): When newly- 

 hatched, about 1mm. in length, rather stout in proportion ; the colour 

 pale ochreous-green, darkest on the dorsal surface, becoming paler 

 laterally, and palest on the venter ; the body sprinkled with dark 

 brown warty spots, some extremely small ; the largest, those on 

 dorsal area, emit long greyish hairs curving backwards, excepting 

 those on prothorax, which curve forwards ; the other warts ('? primary 

 tubercles) principally run in longitudinal rows, forming supra- and 

 subspiracular series, each emitting a fine greyish hair. Head black ; 

 the true legs brown ; the prolegs of the same colour as the ventral 

 surface. First moult end of April. Second instar (May 1st): About 

 2mm. in length ; the back arched; the sides very sloping; the ventral 

 surface much flattened. Head shining black, retractile within, over- 

 lapping prothorax ; colour of body pale ochreous, with mediodorsal, 

 subdorsal, and lateral, dark- cbocolate-brown, longitudinal lines, each 

 bordered by a whitish-buff line ; the whole surface sprinkled with 

 dark brown, watery-looking discs, flattened and semitransparent on 

 their summits; there are also longitudinal rows of greyish -green, 

 short tubercles, each emitting a curved whitish hair. On the pro- 

 thorax, a dark purplish -brown dorsal blotch. The true legs black ; 

 the prolegs ochreous. Third instar:* 6'25mm. in length; colour pale 

 greyish-green, with a dark chocolate-brown mediodorsal stripe, 

 bordered on either side by a white lint', and a double grey-green 

 (darker than ground colour) subdorsal stripe composed of two slightly 

 oblique marks on each segment, each bordered above by whitish ; a 

 lateral dilated white stripe, bordered below by an olive band ; the 



Chapman has carefully proved that the larva of this species has only " four" 

 stadia, and three moults. Froliawk calls this the " Fourth " instar, and notes the 

 "Third" as exactly like the second, except in size, length = 4mm. Greal care 

 must always be exercised in accepting Frohawk's statements as to the stadia in 

 which larva 1 areas he appears I mm his descriptions, not to follow up any particular 

 arva, or to give data of invariable structural items, e.g., size of head, etc., in a 

 given instar, hut guesses the stadium by the size of the larva, sometimes, as in this 

 case, apparently, erroneously. 



