190i.] , , PJiOM IIHITISH EAST AFRICA. 343 



to ])e coiapai'ed (ugether. Up to the present tbe.se and P. violacean 

 are the .only throe species o£ the genus in which they have been 

 discovered. i have re-examined P. gregorlanus and P. fiani in 

 the hope of finding spermatophores, but quite in vain. The 

 matter \^'ould be ot ob^■io^s interest, since it is clear from tlie 

 dravvings exhibited that the t\\'o species with which I am con- 

 cerned in the present communication can be amply discriminated 

 by their spermatophores alone. In P. magilensis they are all of 

 the same form and vary only in their length, Avhich is in some 

 cases (see text-fig. 85, p. 342) considerably in excess of the length 

 which I originally figured. The spermatophore is extremely thin 

 in proportion to its length, and is expanded invariably at one end 

 into a spoon-shaped head. It dies away at the other end to a 

 fine point. The body throughout is closely beset with a fine 

 covering of spermatozoa, and I cannot but doubt that these 

 spermatophores also are mobile during life. But they were not 

 twisted into the tight coils so characteristic of those of P. Jdndei. 

 The relati\'e sizes and the dilferences in structure can be gathered 

 from the sketches (text-figs. 84 & 85), 



In originally describing these spermatophores I commented upon 

 their likeness to immature spermatophores^in Tubifex, and thought 

 that they might be in a state of immaturity. However, the occur- 

 rence of abundant spermatophores of precisely the same form in a 

 second individual of the same species which I record here seems to 

 me to do away with that possibility. Furthermore, there were no 

 intermediate stages which would suggest a development of the 

 theoretically immature spermatophores of P. magilensis into fully 

 formed ones like those of P. hindei and P. violaceus. All the 

 spermatophores — a very large number and, as I have already stated, 

 from two individuals — were at precisely the same stage, which 

 must be therefore, I should imagine, their definitive stage. 

 Among the fully-formed spermatophores were a number (text- 

 fig. 85, B) which had lost their contents, and were simply empty 

 sacs, more or less hyaline in character, and preserving the exact 

 form of the uninjured spermatophore. This material was not 

 suggestive of the granular matter figured by Vejdovsky in the 

 immature spermatophore of Tuhifex, which he ascertained to 

 proceed from the secretion of the cement-gland or at least from 

 the atrium of that worm. It appears to me, in fact, that the 

 spermatophores of Polytoreutus magilensis are really constructed 

 on the same plan as those of the two other species of the genus 

 in which they occur, but that the actual case of the spermatophore 

 is much more slender and thus the spermatozoa project much 

 further out. The result is an entirely diffei'ent aspect, which is 

 well shown in the drawings exhibited herewith. As to the 

 spermatophores of Polytoreutus violaceus and P. Jdndei, their close 

 likeness to those of the Tubificidse other than Bolhrio neuron is very 

 striking, and applies to details of structure, Yery often, though 

 not in every case, the anterior end of the spermatophore, which is 

 sometimes slightly swollen, was distinctly open, as shown in the 



23* 



