some Dipterous Larv(B, 607 



nor indeed with the Diptera Aschiza in general. For the 

 purpose of comparison, I tabulate below some of the more 

 important differences between Eristaline larvae and these 

 examples from the gill-chambers of crabs : — 



Eristaline larvce. Larvce A and B. 



(1) Siphon not bifurcated, posterior (1) Siphon bifurcated, posterior 



spii-acles adjacent. spiracles separated. 



(2) Protlioracic spiracles of small (2) Prothoracic spiracles well de- 



size, veloped in larvte A, larva B 



is in the second sta^e only. 



(3) Antenna and maxillary palp (3) Antennae and maxillary palp 



united to form a single sen- distinctly separated, 



sory organ. 



(4) Bucco-pharyngeal apparatus (4) Bucco-phaiyngeal apparatus 



exhibiting a special struc- of the ordinary type common 



ture, quite dirterent from the to all C^'clorhapha Schizo- 



type usually found in Diptera phora. 

 Cyclorhapha. 



All this shows that the resemblance between tlie larva? 

 A and B and the Syrphid larva is only a superficial one — it 

 is nothing more than a case of convergence. 



The question now arises: What is the correct systematic 

 position of these larvae ? 



In the existing state of our knowledge of the larvae of 

 Cyclorhaphous Diptera, it is always difficult, and often 

 impossible, to determine even the family to which a parti- 

 cular example belongs. Nevertheless, the structure of the 

 head, mouth-parts, and spiracles indicate, first of all, that 

 the larvae A and B belong to the Diptera Cyclorliapha 

 Schizophora. They appear to show a close affinity with the 

 the family Ephydridae — this for the following reasons : — 



(1) Larvae of the Ephydridae, like larvae A and B, are 

 always provided with postabdominal respiratory siphons. 



(2j In this family the siphon is always bifurcated as in 

 larvae A and B. 



(3) The prothoracic spiracles are well developed in the 

 Saprophagous Ephydrid larvae, similar to those of larva A. 



(4) The mouth-parts of larvae A and B are of the Ephydrid 

 type. 



(5) The ventral surface of the head in Ephydrid larva?, as 

 in larva A, is furnished with plates and dentate scales. 



(6) Several species of Ephydrid larvte are known to occur 

 in salt or brackish water. 



It is, therefore, highly probable that tiie larvae A and B, 

 discovered by Baylis, belong to the family Ephydridae, or, 

 at least, are very closely allied to that family. 



