and rules the exterior form of the caterpillar. The intention of this 

 hairy covering, as I have already said, it is not in my power to explain. 



As to the form of the bristles, a few observations, not without in- 

 terest, may be offered. In the species of Melitcea and Argynnis which 

 I have examined the bristles present in the first stage are always long, 

 slightly curved and finely toothed on the margin. These primitive 

 brisdes, as I will call them, are now, in the course of the development, 

 replaced by bristles which are either smooth and simple, or swollen 

 near the base. This process of replacement is less rapid in some spe- 

 cies than in others, but it becomes effectual either in the later or the 

 earlier stages. Thus, Melitcea Phaeton retains its primitive brisdes to 

 the fourth stage; whether it does so to the fifth I know not, as that 

 stage was not at my disposal. In Marcia and Nydeis, on the con- 

 trary, single brisdes of the second with swollen shaft are seen directly 

 after the first moult. They make their first appearance, mingled with 

 the primitive brisdes, on the warts of the sub-dorsal row. Not until 

 the following stages do they crowd out the primitive bristles on this 

 and other rows. In Melitcea Marcia we find in the second stage a few 

 club-shaped bristles on the sub-dorsal warts, while in all other positions 

 the bristles are toothed; in the third stage, on the contrary, the latter 

 are confined to the infrastigmal row, and even there a clubbed brisde 

 occasionally appears among them. In the fourth stage of Nydeis (the 

 first moult after the hibernadon) the second form of bristles gives place 

 to a third, while the swelling has almost entirely disappeared from all 

 the bristles. 



In the fifth stage of Marcia, also, there would seem to be a tendency 

 to change the swollen bristles for simple staff"-shaped ones. If we 

 designate the three kinds of bristles, viz., the toothed, the swollen and 

 the staff"-shaped, as A, B and C, we may arrive at the following con- 

 clusion: MelitcEa Phaeton is the oldest species, standing nearest to the 

 prototype, because in it the primitive form of bristles A persists until 

 the fourth (perhaps the fifth) stage; Melitcea Marcia is a more recent 

 species, because the form B appears as early as the second stage, while 

 a tendency to form C does not become observable until the fifth. The 

 newest of the three species would be Melitcea Nycteis, because the form 

 C entirely supplants B as early as the fourth stage. In Argyjinis 

 Myrinna the toothed primitive form is also found in the first stage, 

 but it is immediately supplanted after the first moult by the form C, 

 and not by B. 



In the species of Grapta which 1 have examined the bristles of the 

 first stage are not generally toothed, but mostly quite smooth. The 

 warts make a more considerable growth than in Melitcea and Myrinna, 

 while the bristles become proportionately shorter. 



