502 BRITISH LEPiDOPTERA. 



Variation of larva. — Speyer notes (Stett. Ent. Zeit., xxxviii., pp. 

 40 et seq.) that the ground-colour of the full-fed larva varies. Four 

 (out of seven) larva? were pale yellow, one lemon-yellow, and two 

 pale greenish-yellow. The principal markings are the four longi- 

 tudinal rows of black spots, of which only the upper pair are constant, 

 and these are also very variable in size and shape. Four of the larva? 

 examined had both upper rows composed of two spots on each seg- 

 ment, of which the former is the larger, and either square or round, 

 whereas the hind one is hollowed out on its front border, and is nearly 

 half-moon shaped. The other two rows are also composed of two 

 spots on each segment, of which the front one is generally placed 

 rather lower down, and is larger than the hind one. Both are of an 

 irregular, sometimes long, sometimes round shape. On the first two 

 segments the spots are smaller, narrower, and the upper pair very 

 often joined. Of the remaining three larva?, two have the lower rows 

 of spots represented by points, and the other has lost these altogether, 

 with the exception of a single dot on the 4th and 6th segments. In 

 these three larva? the spots of the upper rows are also comparatively 

 small, the front one is square on the middle segment, the back one is 

 represented by two dots, one above the other ; on the first five seg- 

 ments half-moon shaped. The lateral spots of the three most strongly- 

 marked larva? form a black longitudinal streak, or rather a row of 

 spots, as the streak is interrupted by the segments. Of the remainder, 

 two have, instead of the streak, an interrupted grey line, which is 

 shortened in front and behind, and the other two have lost even these 

 markings. In the place of the dorsal plate, two larva? have a black 

 spot on each side, and behind this a lateral stripe ; these larva? have also 

 a broken grey line, reaching from the 4th to the 9th segment, along 

 the middle of the venter. The remaining larva? have not these markings. 

 A small row of bright, yellow, lateral spots, placed between the upper 

 and lower rows of dorsal spots on the hind border of each segment, 

 is, however, quite constant. 



Comparison between the larv;e of A. palustris (trifolii-major), 

 A. trifolii and A. filipendul^ . — We have given at length two de- 

 scriptions of the larva of A. palustris, made in 1871 (Briggs) and 1897 

 (Tutt) respectively, and quite independently of each other, simply to 

 show that they are identical, and because of the important bearing 

 this has on what follows. In comparing our description of the larva 

 of A. palustris (trifolii-major) with that of Buckler's A. filipendulae 

 (Larvae, etc., ii., p. 98), we were astonished to find that the black 

 spots of the dorsal stripes are arranged almost identically with those 

 of the same stripes in A. filipendulae, i.e., large and small, on each seg- 

 ment, and not X-shaped, as in the true A. trifolii, a character which 

 Hellins says is the chief distinction between the larva of this species 

 and that of A. filipendulae, L. (Ibid., p. 96). As our larva was taken 

 from a cocoon that was picked from a piece of marshy ground, and all 

 the other cocoons produced veritable A. palustris, there could be no 

 mistake, and, moreover, the identity of our description with that of 

 Briggs leaves no room even for suspicion. We observe, too, that 

 Briggs has a note on this point which we unhesitatingly confirm. He 

 writes (Trans. Ent. Soc.Lond., 1871, p. 438) : " As the variation in the 

 larva of A. filipendulae tends towards confluence, the variation here is 

 towards obliteration, but the caterpillar is very constant, its range of 



