ANATOMY AND CLASSIFICATION OF THE AEENICOLTDJ!. 553 



The testis of A. ecaudata is a thin, flattened, and slightly 

 folded reniform plate, and its mode of origin is similar to the 

 ovary. The early stages of the spermatozoa are completed in 

 this organ. Ruptnre of the wall sets them free into the body- 

 cavity, but whether the ovary or the testis subsequently 

 undergoes changes by which a new formation of ova or sper- 

 matozoa is accomplished in the following year, or whether 

 periodic discharge takes place at shorter intervals, we have 

 not ascertained. In any case the result of the development of 

 vascular ovaries and testes in A. ecaudata is, that this species 

 retains the gonads within the reproductive organs for a much 

 longer time than is the case with its allies, and discharges 

 them into the coelom in a more fully developed condition. 



In the other species of the genus the gonads are merely 

 proliferations of the peritoneum of the nephrostomial vessel. 

 They are absent from the first nephridia. Each of the follow- 

 ing nephridia bears a single gonidial strand on its inner side. 



In all species all the nephridia act at certain times as the 

 oviducts or sperm-ducts. In A. ecaudata a special diver- 

 ticulum of the vesicle often occurs, in which the ova or sperms 

 are stored for a short time before being expelled. 



XIV. Branch iomal dan e Vincent i resembles in its ex- 

 ternal characters a late post-lai'val stage of one of the ecau- 

 date Arenicolidfe. The internal anatomy differs in the 

 absence of otocysts ; the origin of both ovaries and testes 

 in the same animal from the septa ; the variable number of 

 nephridia (two to five pairs in segments 5 — 9). The resem- 

 blances lead us to include it in the family Arenicolidae, but 

 in a distinct genus. 



XV. Though the Arenicolidae show a closer affinity to the 

 Maid anida3 than to any other fainily of Polychaetes, they form 

 a very well-defined group. The points of resemblance to the 

 Maldanidfe are limited to the parapodia, the prostomium, 

 the brain, nerve-cord (especially the structure and arrange- 

 ment of giant-cells in the cord), the nephridia, and gonads. 

 The differences, however, are quite as striking, e. g. the seden- 

 tary, strictly tubicolous habit of the Maldanid^, their simple 



