33 



course of 21 hours; in my own experiments I did not lind I In- young 

 Plutei until about two days alter fertilization. 



As described by Tenncnt this lirsl larval stage (I'l. VIII, Fig. 2) has 

 the posterior end of the body truncated, the body skeleton forming a very 

 regular basket-structure. The skeleton (Fig. 6) is very smooth, the body 

 rods and ventral transverse rods, as well as the rods connecting the re- 

 current rods are slightly widened at the end, where they join. There is 

 a pronounced and characteristic sinuation at the point of issue of the 

 posloral rod from the body rod. The posloral rod, which is gracefully 

 curved at the base, is fenestrated; but the holes are small, excepting the 

 lowermost one, and gradually disappear towards the point. Red pig- 

 ment is beginning to appear in the point of the posloral arms, and also 

 some few scattered red pigment cells are found in the body. 



In this stage the larva 1 remain essentially unchanged for an unusually 

 long time. It was not until the 121 li day after fertilization that the form- 

 ation of the posterodorsal arms began. That this long duration of the 

 first larval stage is a normal feature can hardly be doubled, since the 

 larva? afterwards developed quite normally. Also two other cultures gave 

 similar results, though these latter were not kept so long a time that 

 the beginning of the second stage was reached. 



At the age of 17 days the body skeleton had begun to be absorbed, 

 and the larva was about to assume its full shape (PI. II. Fig. 1 ); the vibralile 

 band had reached the hindend of the body on the sides, being drawn out 

 into a pair of lobes, the poslerolaleral processes. The posterior trans- 

 verse rod was found to appear on the 201 h day. 



The fully formed larva (PI. II, Fig.2), 22 days old, is a strikingly beautiful 

 object. The arms are all broad, especially Ihe anlero-lateral arm is 

 broadly widened at its base; the posloral and the poslero-dorsal arms are 

 slightly and gracefully bent. The arms are all rather short. Ihe longest, 

 the postoral ones, only slightly exceeding the length of the body. The 

 poslero-dorsal arms are somewhal shorter, and Ihe preoral and antero- 

 lateral arms are generally quite short, though they vary somewhal in 

 length. The poslero-laleral processes remain short, ear-shaped lobes: 

 above these lobes I In- body is distinctly narrowed, which feature, together 

 with the concave outline of the posterior end of the body, gives Ihe fully 

 formed larva a very dilTerenl appearance from thai at the beginning 

 of Ihe transformation from the lirsl to the second larval slage (Comp. 

 PI. II, Figs. 1 and Fig. 2). There are two pairs of vibralile lobes (auri- 

 cules); the dorsal ones, a I Ihe base of the poslero-dorsal arms, are out- 

 wards directed, the ventral ones, between Ihe base of the posloral arms 

 and the anal lobe, are forwards directed. The anal lobe is deeply concave 



5 



