THE EARLY STAGES OF HBSPfJRIA LINEA. 255 



surface and on the anal extremity. The surface is roughly 

 granulated, and of a pale citron-yellow colour. 



The larva eats almost all the shell after emerging, only 

 leaving a portion of the base adhering to the surface of the 

 sheath. Shortly after emergence the young larva commences 

 spinning itself over with silk until it is completely enveloped in 

 a little dense, elongated oval, white cocoon, spun in the same 

 spot where the egg is laid, so that a row of cocoons takes the 

 place of the eggs, the little larvae not moving from where they 

 hatched. In these compact cocoons the larvae are completely 

 hidden and protected for hibernation. In this respect, except- 

 ing its near ally, H. action, which I shall later on fully describe, 

 this very remarkable habit of H. linea is unique among 

 our native butterflies, and is a wonderful provision for the 

 preservation of the species ; as the eggs are but slightly 

 fixed to the surface of the sheath on which they are laid, 

 they become detached if only very lightly touched, and upon 

 the decaying of the grass during winter, when it would become 

 fractured and split up, the eggs would fall out and perish on the 

 ground in consequence ; but Nature has provided safety for the 

 species by the self-preservation of the larva to construct an 

 indestructible covering incapable of dislodgment from its site, in 

 which it can safely remain concealed throughout the winter 

 months, and in the spring, when the fresh, tender blades of the 

 grass-plant spring up among the fallen flower-stalks and sheaths, 

 the little larva, on awakening, can then find its natural food — at 

 least, this appears to me to be the obvious reason of the necessity 

 of this wonderful provision. 



In Buckler's 'Larvae of British Butterflies,' pp. 195-6, Hellins 

 states that on July 29th, 1865, Dr. Knaggs sent him *' the eggs 

 of this species laid in a row in folded grass, but how he managed 

 to get the butterfly to lay them I do not know. The larvae 

 hatched on August 12th ; most of them soon disappeared, but 

 one survived until the middle of November ; it was then only 

 about 2 mm. in length, so I must have mismanaged it." Also 

 he mentions being unable to obtain a single egg from numbers 

 of these butterflies in captivity ; but the first egg he squeezed 

 from each dead butterfly proved fertile, and " the larvae began to 

 hatch August 15th ; they soon spun little ropes of silk across 

 the blades of grass, and made little web coverings for them- 

 selves, but they would not feed, and an accident soon befel their 

 cage, and I saw them no more." 



Both from Hellins' description of the egg and his account of 

 the site of egg and habits of the young larvae, it is not at all 

 conclusive if he had under observation H. linea at all, although 

 it appears impossible that he could be mistaken in the identity 

 of the specimen he squeezed eggs from ; but one point is, I 

 think, certain, viz. he described the egg of H. action for that of 



X 2 



