L84 



Part II T. — Fifteenth Annual Report 



that the larvae could be reared successfully on dead matters of a more or 

 less artificial origin. 



New forms of food were occasionally observed ; of larval molluscs, I 

 have figured two specimens, which to all appearance were larval Gastropods 

 (figs. IS and 19). A new crustacean was also now seen in the stomachs 

 (fig. 20), but from the best preserved specimen which is figured, it is 

 difficult to recognize the species. There is reason to think it is also a 

 larval form. 



The more rapid progress in development that was observed in the pre- 

 vious period was continued during this, and, if anything, increased. 



An average sample measured 9*64 mm. in length, and 2-16 mm. in 

 breadth. The latter increase shows how the larvae were now gradually 

 getting broader compared with the length (fig. 3). 



The posterior extremity of the spine had now begun to decrease in size 

 and bend more upwards (dorsally), while the ventral portions of the tail 

 had been much broadened. 



The number of cordal spines had much increased, and dorsal and 

 ventral fin-rays were present. The position of the stomach was changed, 

 as the sling of the mid-gut had developed more in dorsal and ventral direc- 

 tion than posteriorly. The anus was therefore now situated relatively 

 more anterior than in the earlier stages. 



The left eye appeared at this time in a more anterior and dorsal posi- 

 tion than the right, but the larvae showed no other sign of transformation. 

 The pigmentation was of a fainter appearance than in the previous stage ; a 

 large number of brick-red spots were scattered evenly over the body and 

 mixed with the black pigment, but it had not appeared on the fins. The 

 black pigment alone extended over the caudal fin. 



mh to 4:0th day. 



I have mentioned already how deposit from the tow-nettings had 

 accumulated at the bottom of the jar, and that the larval plaice found 

 abundant food there. By this time young jellyfish occurred in the tow- 

 nettings, and some had been allowed into the jar. They ultimately died, 

 and sank to the bottom where they decomposed. The consequence 

 was that some larvae, in one way or another, got such matters amongst 

 their food and were quickly killed ; I lost several good specimens in this 

 way before this deposit could be removed. The Evadne continued as 

 one of the principal forms of food ; in fact, I never found a larvae with- 

 out it at this time. In the gut of one specimen I counted remains of five, 

 while the stomach itself contained many more. Such was also the case 

 with Scrobicularia alba, though not quite to the same extent. Additional 

 forms, however, were often seen. I have figured another larval crusta- 

 cean (fig. 23) which has not been identified. Of recognisable forms I now 

 commonly found amongst the food the following copepoda : — Temora 

 longicornis (Miiller), Centropages hamatus i Acartia longiremis (Lillje- 

 borg). At the end of this period the measurements along a central line 

 were as follows on an average-sized specimen (fig. 4) : — 



From the tip of the snout to the clavicular bones, . . 2*32 mm. 

 From the clavicular to a vertical line touching the posterior 



curve of the gut, 1*92 mm. 



From this line to the root of the tail, . , . .5*20 mm. 

 The length of the tail, ... .... 1*52 mm. 



Total length of larva?, . , , , , . 10*96 mm. 



