DECAPODA. 



207 



the parva stage onwards. He did not possess any younger specimens, but suggested 

 that the earliest form was in all probability a zoaea. The material in my hands enables 

 me to confirm this view. In the case of Acanthephjra {Systellaspis) debills, a species 

 of which the development is now almost completely known, the larva is born with the 

 pereiopods fully formed and the pleopods present as buds ; in this instance, however, 

 the ova are relatively of a much greater size than is the case with A. purpurea. 



The following table shows some of the features of the specimens examined, excluding 

 those which were too broken to afford accurate data. The rostral length is subject to 

 considerable variation during the different stages : I have therefore measured the 

 specimens from the blunt tooth, which is situated at the base of the rostrum, to the tip 

 of the telson (excluding setae) ; in the younger forms this tooth is a very prominent 

 characteristic, and it appears to persist throughout the stages as the posterior dorsal 

 spine of the rostrum. 



The figures in the second column can only be regarded as approximate, for the accurate measurements of 

 specimens possessing varying degrees of curvature is necessarily attended with some difficulty. 



The smallest specimen in the collection measures about 43 mm., and there can be no 

 doubt that the larva is but little smaller than this when it leaves the q^^. 



As will be seen from the table, the collection contains nineteen specimens, ranging 

 from 4-3 to 5'2 mm., in which neither uropods nor pleopods are developed; these repre- 

 sent the earliest stage* in the metamorphosis (PI. 14. tigs. 1-8, 11, 14, and 18). 



* It will of course be understood that the word " stage " is merely used to distinguish larval groups differentiated 



35* 



