558 EEPORT OF COMMISSIONER OF FISH AND FISHERIES. [104] 



as in protozoa and amcEboid forms generally the food particles are taken 

 into spherical cavities and surrounded with a watery fluid to which fer- 

 ments are probably added during digestion from the surrounding 

 plasma, so as to dissolve and incorporate by difiusiou and intussuscep- 

 tion that which was not into that which so becomes part of the amoeba. 

 The presence of clots of amorphous matter in the cavities of fish embryos 

 when sections are prepared has often struck me as evidence of some- 

 thing of the sort described above. These may, however, be artificial 

 products, and the effect of the extractive and solvent action of chromic 

 acid and the precipitating action of alcohol. 



Not less interesting than the phenomena just described are some of 

 the irregularities of cleavage. These irregularities were formerly not 

 much noticed, or if noticed, investigators were not in a position to 

 assign to them their true significance. It is probably true that most 

 embryos will be found to vary more or less notably, if the segmentation 

 of large numbers of ova be carefully studied and compared. We know, 

 for example, that two well-marked types of segmentation arc found to 

 obtain in Ostrea virginica, as shown by Brooks. Those of other types 

 also are known to vary, often greatl}', in the details of the relations of 

 the segmentation furrows, especially at their points of meeting — that is, 

 where new furrows run towards and join preceding ones. From the 

 investigations of Rauber, we may infer that in some forms this irregu- 

 larity is very marked — so much so, that outline diagrams of the cleav- 

 age furrows of, say, the morula stage of any two ova of the same species 

 would nowhere exactly coincide if superimposed. This irregularity is 

 found to obtain extensively in meroblastic ova, and is, perhaps, almost 

 constant in them, so that we discover that a tendency toward individual 

 variation exists in part at the very beginning of development, without, 

 however, interfering with a well-marked or characteristic plan of devel- 

 opment in the case of each form. 



The irregularities in the cleavage of the germinal disk of fishes become 

 evident at a very early stage, but most conspicuously after the germ has 

 been divided into four cells, as shown m Figs. 33a, 336, and 39. Slight 

 differences in size may also be noticed at the time the germ is divided 

 into two segments, as in Fig. 44. In Fig, 33&, giving an outline repre- 

 sentation of the germ of the shad's egg, the whole has an oblong, sub- 

 quadrate form, as seen from above. The first transverse furrow ii, a, 

 ii, is bent obtusely at the two points where the furrows of the second 

 cleavage i, i, meet it. This is actually an effect of the second segment- 

 ation, because we do not find a short section a of the first cleavage fur- 

 row having a course different from that of the outer limbs ii ii prior to 

 the second cleavage. At the time the first segmentation furrow is fully 

 formed in fish ova, generally, it divides the germ disk straight across, 

 as in Figs. 9, 35, 37, and 44. The inequalities observable in the arrange- 

 ment of the furrows seem to be mainly due to the displacement of the 

 cleavage planes, or, perhaps, more properly, to a slight angular shifting 



