924 LORD WALSINGHAM ON THE [Nov. 26, 



I met with a single small larva (9841 7) at Guimar, on March 1 1th, 

 on Phagnalon saxatile, the flowers of which it continued to eat 

 very sparingly until it died on May 25th : persistent efforts to 

 Jind other specimens wei"e unsuccessful. The brownish larva, 

 less than | inch in length, has a pair of small projecting pronotal 

 tubercules, and a metanotal pair, larger and more erect, also an 

 anal tubercule ; on all the segments are short, reflexed, bristles, 

 arising from pairs of small tubercular excrescences. It differs 

 somewhat from the larva of the allied saianas Mill., and I had 

 regarded it as probably that of adactyla Hb., with which I am 

 not personally acquainted, until observing the close alliance of 

 the imago of canariensis Rbl. to that species. There can be 

 little doubt that this was the larva of canariensis, although I so 

 unfortunately failed to rear it. 



(1426) AoDtSTIS SATANAS Mill. 



Agdistis satanas Mill. Bull. Soc. Ent. Fr. XLIV. (5 s. V : 1875). p. clxvii (1875) ' ; 

 Cat. Lp. Alp-Mar. 377-8. PI. 2 • 9 (1875) 2 : Nat. Sic. V. 221-2 no. 3114 ^^^ (1886) » : 

 Stgr-Rbl. Cat. Lp. Pal. II. 78 no. 1426 (1901)^; Wlsm. Ent. Rec. XIX. 53 (1907) '=. 

 Sab. S. France ^'^ -. Cannes, ® Scabiosa candicans VI, excl. VII ■*. 

 In Mr. Tutt's recently published ' British Lepidoptera ' [V. 129, 136 (1906)] some 

 doubt is expressed as to whether a larva which I submitted to Dr. Chapman as that 

 of Agdistis satanas Mill, is reallj^ an Agdistis at all : this opportunity maybe taken 

 to record the evid'ence upon which the identification rests. My experience in attempt- 

 ing to rear canariensis was precisely similar to that of Milliere, and the failure may 

 probably be attributed to the hibernating habits of the larvae of this group coming 

 into premature practice through the necessity of keeping the bottles containing them 

 in a comparatively cool temperature. I extract the following abridged notes from 

 my voluminous correspondence with my late friend Monsieur Milliere: — In 1885 

 Milliere had two larvae which he believed to be those of Agdistis satanas. He was 

 taking great care of them — one disappeared, the other fixed itself up for pupation, 

 but did not change, drying up, but preserving its form, so that he could figure it. 

 In his letter to me, of August 19th, 1885, he mentions this fact and adds "car je crois 

 avoir acquis la preuve que c'est bien la chenille de Satanas." The proof appears in 

 the following translation : ' One female of this Agdistis has had the good thought 

 to lay ten fertile eggs, but I have not been able to feed the little caterpillars, which 

 have not touched anything and have died of hunger. I have preserved in spirit 

 some of these young larvae, which, under a strong lens, seem to me to have all the 

 characters of the caterpillar which you have prepared for your collection. I can send 

 you these larvae obtained ab ovo.' I have empty egg-shells, sent at that time, but 

 cannot find the larvae in spirit. 



In June 1886 Milliere published (Nat. Sic. V. 221-2) the following additional 



information on Agdistis satanas : — 



" 06s. Au dernier moment je trouve a I'habitat de la Satanas, la chenille de cette 



Agdistis qui, du 15 au 25 juin, est parvenue a son entier developpement. Elle se 



nourrit sur la Scabiosa candicans dont elle ronge les feuilles, et sans doute sur 



d'autres plautes sous-ligneuses. 



" Cette larve rappelle la chenille de sa congenere Heydenii, mais elle est plus courte, 



avec les caroncules dorsales moins developpes et les polls longs et raides, dont elle est 



couverte, depourvus, a I'extremite, du petit rentlement spatulifere qu'on remarque 



chez sa voisine. 



" L'^tat 16thargique dure a peine un mois. 



" L'Agd. Satanas qui n'a qu'une g<5neration pond rarement en captivite, cependent 



une '^ enfermee en un tube de verre, ay ant pondu une 30* d'oeufs, j'ai pu les etudier. 



Ces oeufs sont relativement gros, elliptiques, blanchatres, et profondement canneles 



en long ; leur eclosion eut lieu 15 jours apres. 



"La jeune ch. se montre alors a pen pres ce qu'elle sera a ses divers ages. Elle 



passe I'hiver fixt^fl a une tige seche, dissimulee dans les brindilles herbacees." 



