THE SILK-SPINNING HABIT IN BUTTERFLY LARVE. 59 



bridge was built over the torrent, and upon it, lay motionless, and out 

 of danger, the little larva." 



The larvae of Epinephele ianira, and other Satyrids, spin together 

 grass leaves for a puparium. Polyommatus icarus, Lampides boetica, 

 and other Lyca3nids, will similarly spin together leaves and flowers of 

 their foodplants. Chapman says that the larva of L. boetica will fasten 

 together flowers into a cocoon for pupation, or pupate openly on a pad 

 spun on a flat surface, and asks — Is the puparium the result of 

 spinning a carpet amongst loose material ? or, does the carpet result 

 from making the puparium on a flat surface ? Poulton considers 

 {Trans. Ent. Soc. Lond., 1904, pp. cxii-cxiii) that the silken material 

 spun by butterfly larvae for suspension, either by the anal segment 

 alone, or by a girth in addition, is, in all probablity, the persistent 

 trace of a vanished cocoon, and he surmises that the decline of the 

 cocoon, a form of passive defence, built to endure for comparatively 

 long periods, including the time of special stress, was favoured by a 

 short pupal period falling wholly within the time of least stress, and 

 suggests that, when the cryptic colouring of the bare pupal surface is 

 as effective for concealment as that of the cocoon, it presents certain 

 advantages over the latter, etc. The change from a hidden to a 

 cryptically protected pupa, would gradually lessen the need of a 

 specially prepared cocoon, and the transition, he says, is easy, from a 

 loose and open cocoon with apertures through which the cryptic 

 colours of the enclosed pupa could play their parts in defence, through 

 stages in which the latter element becomes more and more important 

 as the cocoon progressively diminishes, to the climax when the almost 

 invisible remnants of the silken covering are retained as supporting 

 structures merely. 



The old authors, Swammerdam and Eeaumur, gave very lucid 

 accounts of the silk-spinning done by certain larvae producing 

 suspended pupae — Vanessa io, etc., and those producing pupae attached 

 both by the cremaster and by a silken girth — Pieris brassicae, etc., and 

 the silk-spinning operations performed by such must have been 

 observed by every field lepidopterist, but the details given by Chapman 

 (Trans. Ent. Soc. Lond., 1905, pp. 203 et seq.) are exceedingly 

 interesting. He notes that the first spinning done by the fullfed 

 larva of Thais polyxena var. cassandra, is to form what must be called 

 a cocoon, though it consists merely of three or four, or at most a 

 dozen, rather strong silken cables, sometimes simple, sometimes 

 branching, tying together the objects surrounding the position chosen 

 for suspension. Having prepared a carpet of silk of rather more than 

 its own length, either on a flat surface, or, by preference, on a round 

 one, such as a stem, it makes the anal pad, a somewhat flocculent 

 little mass, and it may be noted here that, in Thais, as well as in 

 Papilio machaon and Pieris rapae, when this is completed, the larva takes 

 its station with the anal claspers just in front of it, the little mound of 

 silk forming the pad being unused, and lying immediately behind the 

 anal claspers and beneath the tip of the anal plate, suggesting that the 

 pupa shall have a freer access to it, than if the anal prolegs held it. 

 In suspended pupae this pad is held by the anal claspers, whilst the 

 larva awaits pupation. The girth arises well forwards from the carpet 

 of silk in Thais, and, as in Papilio and Pieris, is spun in a position 

 that may be described as being in front of the larva, the head being 



