78 PRELIMINAKY EXPERIMENTS ON 



two larvffi were observed to be somewhat feeble, in consequence of the 

 turbidity of the water, and on the ninth day only nine larvse survived. 

 These, however, after careful cleansing of the jar, were all flourishing. 

 They were growing in bulk, and displayed the greatest activity in their 

 search for food. Their numbers remained unchanged for another week, 

 at the end of which time (the sixteenth day) it was clear that the 

 conditions of the experiment were perfectly suitable for their develop- 

 ment, so long as the water could be kept clean and abundance of food 

 could be provided. During the course of the experiment, from one- 

 half to two-thirds of the water had been changed every day, and fresh 

 plankton added daily with rare exceptions. But irregularities in the 

 quantity and condition of the available plankton were unavoidable, 

 causing occasionally a slight deficiency in the amount provided, and at 

 other times an over-accumulation of dead Copepods, etc., in the jar, and 

 consequent soiling of the water. The larvae at this time had grown 

 considerably, being about 7'5 mm. in length, and nearly 2*0 mm. in 

 maximum depth, as taken through the hinder part of the eye. The 

 larval fin membrane was still continuous, but the notochord was turned 

 up posteriorly, the tail fin properly formed, the hypural plates and 

 caudal fin-rays established, the fin-rays in the dorsal and ventral fins in 

 process of formation, and the rudiments of the pelvic fins present. 

 There was thus no doubt that the larvae were growing and developing 

 properly, and were already undergoing the critical stages of their 

 metamorphosis. 



On the other hand, the increasing difficulty of keeping their jar 

 clean in consequence of the large amount of food they required, and 

 the death of a larva on the seventeenth day, induced me to remove 

 some of the larvoB from this jar, and see the effect of placing them in a 

 circulation of the ordinary tank water, which would at any rate solve 

 the difficulty of keeping the water clean. 



Three of the larvK were therefore transferred on the eighteenth day 

 to a portable slate tank, with a glass front, holding four gallons of 

 water, and placed in the main laboratory on the south side of the 

 central tanks. The bottom was covered with gravel, and a circulation 

 was set up with every precaution to avoid loss or injury of the larvae 

 from the current established. On the following day all three were 

 alive and active, whereas another death had occurred in the original 

 plunger-jar. The four survivors were thereupon cautiously transferred 

 to the portable tank, and a liberal supply of fine plankton was provided. 

 Late in the afternoon only three or four out of the seven were 

 swimming about, the others being on the gravel at the bottom. They 

 appeared to be incommoded by the brighter light of their new situation, 

 so a sheet of green glass was placed in front of the tank. No obvious 



