CELASTR1NA ARGIOLUS. 429 



than those of the second brood on the ivy-umbels, the base of the 

 calyx being chosen in most cases. Adkin observes that, in the Hastings 

 and St. Leonards' districts, the 2 of the summer brood (August 

 17th-19th, 1896) selects one of the umbels of ivy-bloom, now quite 

 small, many yet forming compact green heads, closely resembling an 

 unripe blackberry in shape, settles upon its top, closes her wings over 

 her back, and, bending her abdomen down and underneath the buds, 

 affixes an egg to the underside of one of the slender single-bud stalks, 

 at a point about one-third of its length from the crown of the main 

 stalk whence the bud-stalks spring ; but, occasionally, a more advanced 

 head is selected, the buds of which have separated from each other, 

 and the total circumference of which is consequently increased, so that 

 the butterfly, resting on the top of the head of buds, cannot reach the 

 stalks of the lowest row with her abdomen, but the position on the 

 stalk, rather than the particular stalk, appears to be the object she has 

 in view, for she then thrusts her abdomen between the buds and thus 

 attains the desired position on the stalk for depositing the egg. 

 Generally one egg only is laid on a head of buds, but, occasionally, two 

 or even three or four are so deposited. Eaynor further notes that he 

 has once or twice reared C. argiolus from eggs deposited by a $ sleeved 

 on'Rhamnus frangula, whilst Fletcher states that the larvae are to be 

 found on Worthing Downs feeding on the umbels of Cornus sanguinea 

 or may be beaten therefrom, and, on which the eggs, therefore, must 

 have been. laid. [For further notes on " Egglaying" see antea p. 408.] 

 Egg. — The egg is 0*33mm. high, 0-65mm. in diameter. It is flat 

 top and bottom, but the top rounds off into the side rather early, i.e., 

 the actually flat portion is small (about 035mm. across), and the egg 

 makes rather nearer an approach to a bun-shape (as in the Theclid or 

 Chrysophanid eggs), than do most Lycaanid eggs. The flat portion 

 appears also somewhat readily to sink in (on losing mositure), forming a 

 wide, but very shallow, hollow, in the centre of which is the small 

 micropylar depression, as usual free from the white adventitious coat, 

 and only O06mm. across. The white coating is, as usual, arranged in 

 cells, with narrow walls and raised pillars or mounds at the angles. 

 The cells are, for the most part, quadrangular, nearly square, with 

 angles radial and circumferential, and are placed in rows spirally 

 diverging from the centre (engine-turning) with fewer irregularities 

 and deviations than are frequently seen in the pattern of Lycaenid eggs. 

 Bound the sides, however, where the curvature is greatest, there is 

 much variation in the form of the cells ; here and there are diagonal 

 rows of square cells much as on the top; in places, the more elegant 

 (and usual) arrangement of triangular cells arranged in hexagons and 

 pentagons prevails ; here and there the septa run in long continuous 

 diagonal lines, those crossing them being less regular. Sometimes no 

 special disposition can be described, but the cells are always of about the 

 same size and have a most pleasing effect, being, in fact, disposed in 

 apparently every possible way that will give the essential structure on 

 a curved surface. Here and there, however, a pillar seems in reality 

 to be two, not completely fused, a difficulty of arrangement not com- 

 pletely surmounted. Broadly, the arrangement is quadrangular on the 

 upper surface, triangles on the sides, but with quadrangles interspersed 

 to make the fitting easier. As usual the pillars are largest and highest 

 at the top of the sides, where they curve over on to the top of the egg, and 

 become gradually smaller towards the centre, where they are still very 



