184 Proceedings of the Royal Physical Society. 



the rays is such as to indicate that the Solasters have originated from a five- 

 rayed form. The hydrocoele at an early stage forms an open crescent with 

 five primary radial pouches. Then a sixth pouch is added on the sinistral 

 side of V. After that a seventh is added to the sinistral side of VI., and so 

 on till the full number is completed, the very latest being adjacent to 

 ray I., and on the dextral side of the latter. 



It will be seen, accordingly, that the anus in Solaster occupies the same 

 relative position as in other starfish, i.e., it comes after ray V., only here 

 owing to the interposition of new rays, its actual interradius is not V./I. 

 but V./ VI. 



The principle that a serial succession of parts is liable to variation (9) 

 in the number of its components may help to explain the very great 

 variability of both species (and particularly of Solaster papposa (10) ) in the 

 number of their arms. 



3. The Egg Ducts and their Openings. 



As is well known, in most starfish each gonad is provided with a single 

 duct having a single opening on the surface, but in the genus Asterias it 

 would seem that while the duct is single the number of external openings 

 becomes multiplied (Ludwig in Bronn's Klassen u. Ordnungen, Bd. 2, Abth. 3, 

 Buch. ii. p. 597). 



In Solaster papposa (Orossaster papposa, Mull. u. Trosch.), according to 

 Muller and Troschel (quoted from Ludwig, loc. cit. p. 59), the genital openings 

 are also numerous and sieve-like. This was confirmed by Cuenot from the 

 examination of a very large specimen (3, p. 623), though in his former work 

 he had stated that the orifice was single. No account is given by the authors 

 named of the condition of the genital ducts. 



In the closely allied Solaster endeca (Forbes), it is difficult by macroscopic 

 examination alone to determine with exactitude the number and relations of 

 the egg ducts and their external openings. At the Millport Marine Station, 

 specimens have been observed, by Mr Elmhirst the superintendent, in 

 spawning to extrude as many as four ova at a time from a single interradius. 

 This means that each gonad had at least two external openings. Eecently, 

 while working at the growth of the Solaster egg, I made serial sections 

 of the middle portions of two interradii in a large adult, and have 

 accordingly been able to study the points under discussion in the case of 

 four gonads. 



Of these four gonads three had two external openings, while the 

 remaining one had three. The external openings formed slit-like pits, and 



