FISHES. - In the class of fishes, the sper matozoa occur in two forms. The first is 'found throughout the osseous fishes, and also in Amphioxus. The other form is found among the Plagiostomes. In the former case the spermatozoa consist (fig. 346.) of a very small globular body (of -5-ium-1-;#w"', or even smaller, down to Teuu"', as in Peres fluviatilis), and an extraordinarily thin, hair like tail, which, however, possesses compa ratively a very considerable length. Sometimes the body at the point of insertion of the tail has a small knotty appendix, as in Cobitis (fig. 347.), which gives to it a pear-like shape. The body in some genera is so small that it can hardly be perceived with any distinctness.
This also applies to the spermatozoa of Petromyzon, in which the form of the body is, however, different. In P. marinus* the body is egg-shaped; in P. fluviatilis ( fig. 348.) staff shaped. The length of the body in P. fluviatilis is Th"'.
The spermatozoa among the giostonne fishes are similarly formed • to those of the birds. They are long, filiform, and furnished with an terior cylindrical body. In Scyllium - Canicula the body is stiff and quite straight, and tapers at both ends. The tail is thin, and of an equal length to the body (11u"/).
the body, instead of being straight, describes two long spiral windings. Four more narrow spiral windings are found round the body- of the spermatozoa in Spinera acanthias, which measures ,16."', whilst the length of the whole spermatozoon amounts to A similar num ber of spiral twistings are likewise seen in the body of the spermatozoa of most of the rays, in Torpedo narce (fig. 319. B), Raja rubus, &c. In Raja oxyrhynchus it is only the anterior part of the body which is spirally wound in a length of about 1-315"/, whilst the posterior part is straight. The number of the windings is nevertheless, however, rnore considerable, viz. 7 or 8. The length of the whole spermatozoon amounts to Ty". Chimera monstrosa like wise exhibits these windings, notwithstanding the comparatively short body (r,;-6."') of its spermatozoa, which have a length of AP'. The number of windings is three.
The developernent of the spermatozoa in fishes has as yet only been observed in the Plagiostomes. It is exactly the same as in frogs and birds,as the statements of Hallmann* lead us to infer. Almost all of the sperma tozoa are united with one another in bundles. According to our researches in Torpedo Narce, the spermatozoa are produced sepa rately in the cells of developement, which sess about the size of thm, and which are enclosed by a lesser or greater number of cyst-like mother cells (fig. 350.). The size of each of these cysts amounts to about Au", wherever the number of the enclosed cells is small ; but in the reverse case it may in crease to -alum. The cells of developement dissolve after the formation of the sperma tozoa, and the latter then get into the in terior of the cyst (fig. 351.). The spiral wind but here, as well as in the Lacerta, &c., the
formation of the spermatozoa only takes place ings of the body seem to be still wanting at this stage, or, at least, not to be perfectly developed. If the number of spermatozoa is only small in one cyst, they never group together into a bundle, whilst this is con stantly occurring in the reverse case (fig.352.).
We will not venture to decide, however, whe ther thi., difference is entirely attributable to the greater number of the cells of develope ment ; and we are the less inclined to do so, as we have already seen, when investigating the spermatozoa of birds, that, even with an equal number of the formations alluded to, the grouping of the spermatozoa in the in terior of the cyst may be different. This much, however, is certain, that the num ber of the enclosed cells is not entirely with out influence. The fact of the fascicles of spermatozoa alvvays coinciding in one cyst with a greater number of the cells of develope ment, seems, at all events, to favour this con jecture.
Previous to the period of procreation, we also find, in the testicles of the osseous fishes, that the cells of developement are enclosed in the interior of larger cells (fig. 353.); subsequently to the destruction of the cyst, and to the consequent independence of the cells of developement. We infer this from the circumstance that we have never seen in thetn any real bundles of spermatozoa. The state ment of Kiilliker that the spermatozoa of Am phioxus develope themselves from little cells (of „1,,,,"1—,1,"'), which lie together in groups of from six to twenty-five, also seems to support the correctness of our conjecture. Each of such groups appears to us to be the brood of a single mother cell. The mother cells themselves, however, are of such a small size, that the formation of such brood in their inte rior is not to be,traced or perceived. It can only be seen that these cells gradually lose their round shape, and that they assume a pear, or spindle-like form. This, unquestionably, is merely the consequence of the endogenous developement of a spermatozoon, which gra dually stretches itself, thereby causing (as in Gallus, Rana, &c.) the change of shape of the external enclosure.
Thus much of the proportions of form, and of the mode of developement of the sperma tozoa among the Vertebrata. We have treated this subject somewhat elaborately, partly be cause the spermatozoa of these animals are those which may be most frequently obtained for observation,—partly also because it is in them that the stages of developement can be better traced and recognised. We have invariably met with in them a common type, not merely in the external shape, but also in the mode of developement of the spermatozoa ; and these are circumstances which will be of importance to us in interpreting the stages of developement of the spermatozoa in the lower animals, in which they are as yet enveloped in great ob scurity.