1903.] METAMORPHOSES OF DECAPOD CRUSTACEANS. 25 



no mention of these structures in the descriptions of any Crangon 

 larvae, though they are certainly present both in C. trisjnnosus 

 and in C. nanus. It seems to me quite possible that they repre- 

 sent the frontal sensory processes which have been found on 

 certain Cirripede nauplii for example, and which have been 

 supposed to be the vestiges of a pair of preoral appendages. 



In the abdomen, the first and second segments are characterised 

 by the lateral expansion of their epimera (PI. V. fig. 1), which 

 makes them appear considerably broader than the succeeding seg- 

 ments. The third segment bears dorsally two strong backward ly 

 directed spines. In the f oiu-th segment these spines are represented 

 by a pair of very small knobs, while in the succeeding segment 

 they are shifted somewhat laterally and have the form of a pair 

 of long, downwardly curved, blunt processes. The tail-plate is of 

 the usual Crangonid type, with seven strong ciliated spines on 

 either side. These spines, however, are somewhat peculiar in 

 having quite bkmt terminations, a feature which persists in the 

 later stages also. 



The first antennae are unusually long — about two-thirds the 

 length of the carapace. The inner flagellum, which is a direct 

 prolongation of the stem, is ciliated and very bluntly pointed. 

 This blunt termination looks at first sight like the i-esult of an 

 injury ; but as all the larvfe were the same in this respect, this 

 condition is clearly normal. The scale of the second antenna 

 (PI. Y. fig. 3) is long and very narrow, with seven internal setse, 

 two setse and a short spine terminally, and two short external setee. 

 The first maxilla (PI. V. fig. 4) has a two-jointed palp, and in other 

 respects this appendage and the first maxilla also conform exactly 

 to the characteristic crangonid type. 



The exopodite of the first maxillipede (PI. VI. fig. 6) has three 

 apical setfe, and two setje on its external edge, differing in this 

 respect from Crangon vulgaris, which has but one external seta. 

 In the endopodite I can detect only three joints, and in the same 

 way in the other maxillipedes I can only make out four. 

 Williamson shows the palp of the first maxillipede in Crangon 

 vulgaris as distinctly divided into four joints, but though there 

 are certainly four small lobes, I cannot trace the sepai-ation of the 

 second segment into two, nor of the third segment in the case 

 of the other two pairs. 



The more advanced larvte of Crangon fasciatus are compara- 

 tively rare in the tow-net collection, but four stages are to be 

 distinguished in my material. That corresponding to stage iv. of 

 C. vulgaris, i. e. with the pleopods represented by short buds, 

 I have not found, though thei-e can be but little doubt that it 

 occurs. 



In the last stage (stage v.) the leng-th of the body is about 

 3*5-3"8 mm., the rostrum extending about '2 mm. beyond the 

 eyes. The form of the rostrvim, which acquires its definitive 

 shape in the second stage, differs from that of the first stage. 



