18 G. O. Sars. 



at first lie close together, but gradually retire to opposite 

 sides, though still connected by a system of extremely 

 delicate fibres, forming together a somewhat fusiform cen- 

 tral area, in which the chromatic substance of the nucleus 

 is received (see PI. 2, fig. 2). The constituent parts of the 

 latter, the chromosomes, seem to be very numerous, but 

 they are so extremely small and delicate as scarcely to be 

 clearly seen, even by the highest magnifying powers I have 

 been able to employ. I have, however, faintly traced them 

 both at the time when they are still lying densely crowded 

 together in the middle, forming a central plane, and when 

 they have been separated into 2 sets, which by the action 

 of the centrosomes are more or less approximated to the 

 latter. At last only a very narrow bridge is seen connecting 

 the 2 centrosomes, and this is soon interrupted in the 

 middle. Then the division of the ovum itself takes place, 

 resulting in the formation of 2 well-defined segments en- 

 closed within the common vitelline membrane, and each 

 having its own centrosome, which soon again begins to 

 divide in the above-described manner (see PI. 1, fig. 4, PI. 

 2, fig. 3). At first sight, these 2 primary segments appear 

 exactly alike; but, on a closer examination, the one is 

 always found to be a little larger than the other, and it 

 may also be observed, that the division of the smaller seg- 

 ment somewhat precedes that of the larger (see PI. 2, fig. 

 3). Though my observations of these early stages of 

 cleavage are only based upon a restricted number of ova, 

 I have been induced to believe, that of the 2 above-named 

 segments the one is ectodermal in character, the other 

 entodermal; in other words, the numerous cells which in 

 the later stages constitute the ectodermal layer, are all de- 

 rived from only one of the 2 primitive segments (the smaller 



