MUSEUM OF COMPAKATIVE ZOOLOGY. 349 



poles of the spindle into two new nuclei, there is formed about each of 

 the latter a deeply staining homogeneous zone. It appears to be pro- 

 duced from the nuclear fluid, which is pressed out toward the poles, and 

 then consumed in the growth of the new nuclei. In no stage of the 

 division did Mayzel observe a radial arrangement of the protoplasmic 

 granules, such as is to be seen in the eggs of Ascaris. 



The second form of nuclear division was seen in the endothelium from 

 the cornea of the frog, and the epithelium from that of the rabbit. This 

 appears like an hour-glass composed of fibres. The body of the cell may 

 either remain unaltered or may be constricted, in which latter case the 

 constriction closely invests the narrow^ part of the nucleus. The latter 

 corresponds to the " Kernstrange " of Biitschli as seen on germ cells of 

 spermatozoa. In the corneal epithelium of the rabbit, bird, etc. there is 

 formed, independently of the gradually disappearing nuclear fibres, a new 

 equatorial partition composed of granules. In the corneal endothelium 

 of the frog, on the other hand, there appears in the equator of the cell 

 between the nuclear fibres a row of small interstices or vacuoles,- which 

 appear to be filled with a cementing substance. It is from the union of 

 the contents of these vacuoles that the partition arises. A simultaneous 

 division of a nucleus into seven, or even into four parts (Eberth), was 

 never observed. All the important phenomena of the process of nuclear 

 division as given above were also seen in fresh preparations of the frog's 

 cornea examined in aqueous humor. 



The observations of Fcettinger ('76, p. 607) on the division of nuclei 

 in epithelial cells of Petromyzon give no evidence of the existence of a 

 fibrous nuclear spindle. 



BiJTSCHLi's ('77", pp. 212-214) studies on the division of cartilage 

 cells were successful only in making it probable that the nucleus of such 

 cells undergoes a fibrous differentiation preliminary to constriction, and 

 were inadequate to establish any close relation between the nuclear di- 

 vision here and in the typical cases of spindle metamorphosis. He there- 

 fore finds himself forced to the conclusion that a direct comparison of the 

 two methods is not for the present attainable, although the similarity of 

 , the process to that which obtains for the secondary nucleus of Infusoria 

 I makes it reasonable to believe that, in the present case, one has to do 

 i with a modification of the primitive method of nuclear division. 

 1 In a subsequent communication Mayzel ('77") extends his observa- 

 ' tions to the eggs and young stages of Triton and Perca. In segmenta- 

 ! tion spheres of the fish egg from 75 /x to 25 />t diameter, the figures of 

 i nuclear division differ from those previously reported for the endothelium 



