182 CEYLON PEARL OYSTER REPORT. 



short (fig. 35). A larva of 3 millims. is in substantial agreement with the one at 

 2 millims. except that the rostrum is relatively a little longer. 



Larva 4 millims. This stage (fig. 36) is distinguished by the appearance of a 

 small spine at the base of the postei'o-lateral spines of the carapace. The rostrum 

 is relatively longer than in the 3-millims. stage, and now bears a single ventral spinule. 

 A single thoracic segment is exposed posterior to the carapace. 



Larva 5 millims. This larva agrees very well with the one at 4 millims., but 

 the rostrum is a little longer and bears two ventral teeth (fig. 37). Two thoracic 

 segments are now exposed behind the carapace. 



Larva 6 millims. At this stage the antero-lateral spines of the carapace make 

 their first appearance. The rostrum continues to increase proportionally in size and 

 now bears four ventral spinules. The posterolateral spines are likewise relatively 

 longer and the uropods are just discernible as buds. 



Larva 7 millims. The carapace (fig. 38) is now fully formed with anterolaterals, 

 posterolaterals, and the spine at the base of the latter, all well developed. The 

 rostrum is as long as the remainder of the carapace from the anterodaterals to the 

 postero-median dorsal spine, and bears fine ventral spinules. The uropods show one 

 spine on the outer edge of the external ramus. 



Later stages only differ from the 7-millim. larva in the continued relative increase 

 in length of the rostrum and postero-lateral spines, to the proportions shown in the 

 larva 9 millims. (fig. 39), which agrees with the still larger larva, 11 millims. long, 

 described above. 



During development the telson gradually becomes broader in proportion to its 

 length (see figs. 36, 39 and 41), but the number of spines varies very little from the 

 numbers given in the description of the large larva above. 



A comparison of the figures here given to illustrate this Lysioerichthus larva with 

 figs. 1, 2a, 2b, 4 and 5 of Claus' memoir (1871) will show that the species dealt 

 with here is very closely allied to, if not identical with, the one Claus had under 

 observation, and that the three supjx)sed species of Erichthoidina described by the 

 latter author under the names E. gracilis, E. armata and E. brevispinosa in all 

 probability represent developmental stages in the life-history of one species only, the 

 development of which, as gleaned from Claus' figures, follows very closely the lines 

 indicated in the Ceylon larvae. The latter are also in all probability identical with 

 the Erichthoidina figured by Brooks in his "Challenger" Monograph, plate xii., 

 figs. 1 and 2. Brooks has expressed the opinion that Claus' E. brevispinosa is a 

 young stage of Gonodactylus, and bases his view on the presence in the latter of a 

 spine at the base of the postero-lateral spine of the carapace which he regards as 

 characteristic of Gonerichthii. The Ceylon series clearly shows that this spine, 

 though not present in the earliest Erichthoidina, is a later development, and thus its 

 presence cannot be regarded as diagnostic of Gonerichthii, but may be present in some 

 Lysioeriehthii as well (see also Hansen, 1895, plate vii., figs. 4a and 5a, where such 



