Dr. T. A. Chapman on Life History of Lycaena arion. 299 



been made, and to which I have contributed my share, to 

 find the larva with D. flava have met with no scintilla of 

 success, pointing very strongly to the larva never in fact 

 associating with that ant. The hills of D. flava are very 

 often covered with thyme, on which, as on other plants of 

 thyme, the $ $ of X. arion impartially lay their eggs. 

 This, no doubt, is why D. flava has always been supposed 

 to be the ant, if there was an ant, affected by L. arion. 

 It is, of course, the fact that nests of M. scabrinodis are 

 common, often as abundant as those of D. flava, but much 

 less conspicuous, and in more than one instance I can con- 

 firm the remark of Frederick Smith, who says in regard to 

 Donisthorpea {Formica) flava (Brit. Mus. Cat. Fossorial 

 Hymenoptera, Formicidae, and Vespidae, p. 16, 1858) : " This 

 species is sometimes found occupying one side of a hillock, 

 whilst Myrmica scabrinodis appropriates the other." 



I append the notes of Mr. Donisthorpe's observations 

 and my own; it may be useful before giving them to 

 shortly state what they demonstrate. 



When the arion larva leaves the thyme and sets out on 

 its travels there is a vague indication that if it comes 

 across the trail of M. scabrinodis, that is, one of its beaten 

 tracks, it accepts it as a road to be taken. At length it 

 meets or is found by an ant of this species (or some other). 

 It may be, however, that this first and other ants pay 

 little attention to it; at length, however, one does. The 

 ant examines it and proceeds much as ants do when milk- 

 ing Lycaenid larvae ; it goes further than this, it leaves it 

 and circles round it, returns, again milks the larva, and 

 may do this several times. At length, by some agreement, 

 apparently on some signal given by the ant, the larva 

 assumes a most extraordinary form, swelling up the thoracic 

 segments at the expense of the others; such a form as I 

 have seen no other larva assume. The ant then seizes it 

 behind the thorax and carries it into her nest. Here the 

 larva associates with the ants, but receives little or no 

 notice from them, is always at a place where the ants 

 thickly surround a mass of brood, and on this brood the 

 L. arion feeds and grows rapidly to a length of 8 to 10 mm. 

 — so rapidly, that it would be full-grown before November, 

 if it went on ; as it is not full-grown in April, it follows that 

 it takes a winter rest about half-grown. 



My notes with remarks they suggest are as follows, but 

 I place first Mr. Donisthorpe's. 



