39 



was, however, imperfectly developed in the upper part; but the lower 

 part was in very good condition and thus gave sufficient information 

 of the more important features of the second stage of this larva. Xo full 

 figure can, however, he given of the fully formed larva, partly because 

 of the said incompleteness of its anterior part, and partly because the 

 conditions (on board a small launch) did not allow me to make more 

 than some free hand sketches from the living specimen. Although the 

 illustrations given here of the larva of this species thus are not very 

 satisfactory ( I regret to have drawn those of the 7 days old larva 

 in a rather small scale ) the essential characters of the larva come 

 out sufficiently clear from them. 



In the first stage, which lasts until about the 7lh day, the posterior 

 cud of the larval body is short, broadly rounded (PI. VII. Fig.")) and 

 presents no essential difference from that of Li/lirh. iHirii'i/tilnx. The upper 

 part of the anal area, with tin- postoral band, is bent somewhat inwards 

 in the middle, outwards at the sides, the beginning formation of the 

 ciliated lobes. The skeleton is of the same structure as in that species. 

 The postoral rod is simple, not feneslraled, and the body skeleton forms 

 no basket-structure. The end of the body rod has some hori/.onlal branches, 

 very variable in shape; otherwise the body rod is smooth. At the point 

 where the body rod and the post oral rod join then- is a rather deep sinua- 

 lion, generally deeper than in the specimen figured. The recurrent rod 

 is generally bent towards the dorsal side with its point. The ventral 

 transverse rods are transgressing with their points as in /.. i><irit't/<ilus. 



About this time, contemporaneously with the appearance of the poslero- 

 dorsal arms with their supporting rods and of the dorsal arch (PI. VIII 

 Figs. .'5 1), the shape of the body becomes more complicated. On the 

 dorsal side a saddle-shaped impression develops at the level of the lower 

 edge of the mouth; the vibralile lobes are growing larger, and very soon 

 epaulets begin to form, vi/.. about the 10 lllh day. The epaulets are 

 formed in the usual way as an outgrowth from the band at the outer 

 corner of each vibralile lobe. 



In the fully developed larva the epaulets are very large and reach to the 

 middle of the body, where they join with that from the opposite side, so 

 that the four epaulets form together a nearly complete band round the 

 middle of the body, open only on the sides (Fig. X). The posterior end 

 of the body is almost straight, with a pair of short poslero-laleral lobes 

 (Fig. 8,1). Regarding the relative length and shape of the arms in the 

 fully developed larva nothing definitely can be said. The skeleton of 

 the body undergoes the usual absorption and a posterior transverse rod 

 is formed; it is short and straight, with a small process in the middle. 



