471 



MING had established what we now term "chroraatolysis", he described 

 the process (p. 165) in many of them, while failing to identify its 

 nature, or its correspondence with the phenomena, recorded by Prof. 

 Flemming concerning the degeneration of certain germ-cells. 



The connection of the megaspheres with blood-formation he found 

 in the well-known knob on the blastoderm of Torpedo. From my 

 own observations on this structure any relationship to blood-formation 

 must be denied. The cells within it appear to me to furnish as typical 

 examples of chromatolysis as one could wish. 



Finally, Rückert noted the practical absence of his megaspheres 

 in the posterior part of the embryo, a fact of much significance to be 

 commented on in the present work. 



Without searching through the vast literature of piscine develop- 

 ment anew, only two authors can be recalled as having definitely 

 referred to the megaspheres in an attempt to interpret them. H. Ernst 

 ZiEGLER ('96, p. 367) endeavours to explain the megaspheres as the 

 accidental products of an allotment of too much yolk during the clea- 

 vage. His exact words are not quoted. His surmise is very wide of 

 the mark and not a revelation at all of their true nature. He also 

 comments upon their frequently large size, their abnormal chromatin- 

 uetwork, amitotic division (without having witnessed it!), and, adding 

 that their fate is unknown, suggests that they degenerate. 



C. K. Hoffmann, recording them in certain layers of the embryo, 

 interprets them as cells, which have wandered at a later period from 

 the yolk to take part in the formation of the embryo. There is not 

 a particle of evidence supporting this view. 



With the reservation rendered necessary by the vagueness of 

 what is understood by a megasphere, I am strongly inclined to identify 

 Rückert's megaspheres as germ-cells, or, at any rate, as the normal 

 or abnormal fore-runners of such. 



Before the formation of the embryo it will, perhaps, be impossible 

 to pick out the germ-cells with absolute certainty merely by their size 

 or by the amount of contained yolk. The twin-character of the nucleus 

 must be the deciding point. So that in the pre-embryonic period for 

 the time being one can only identify as future normal germ-cells those 

 large cells, which possess this peculiarity; and, as already abnormal 

 germ- cells those "megaspheres", which exhibit the multinucleated con- 

 dition, or budding, or chromatolysis. 



When we come to deal with the megaspheres within an embryo, 

 or those within a blastoderm, on which a definite embryo rests, i. e., 

 in skate of 6 mm whose medullary folds have recently closed, we are 



