314 



SCIENCE. 



[N. S. Vol. IX. No, 218. 



Behreus and Samassa describe for the Sal- 

 monidiB, exists iu the egg of the catfish. 



3. After horizoutal cleavage occurs, the 

 lower cells resulting from this division re- 

 tain their continuity with the yolk, as has 

 been described by Kowalewski, Hoffman 

 and Berent (Teleosts) and Dean (Ganoids). 

 These partial cells (merocytes) continue to 

 divide by mitosis both horizontally and ver- 

 tically. In the former case, the upper of 

 the products of division is added to the 

 germ-disc. This process of supplementary 

 cleavage continues until a late segmentation 

 stage, cells being added to the whole lower 

 surface of the germ- disc. 



4. The periblast arises from the resid- 

 ual portion of the merocytes after supple- 

 mentary cleavage has ceased, being thicker 

 under the margin of the germ-disc, but 

 present elsewhere from the beginning. 



5. The periblast is trophic in its function, 

 playing only an indirect part iu cell-forma- 

 tion. Normal mitosis soon gives place to 

 abnormal and this in turn to amitosis. 

 Transitional forms occur. 



6. The subgerminal space (segmentation 

 cavity) does not appear till about time of 

 origin of germ-ring. At close of segmenta- 

 tion, j'olk and blastodisc are in close con- 

 tact in well-preserved specimens, although 

 no longer continuous with one another. 

 Clefts which early appear between blasto- 

 meres or below them are probablj' artifacts. 

 If not, they disappear later. 



7. The germ-ring (mesentoderm) arises 

 primarily as a marginal ingrowth due to 

 cell-proliferation from germ-wall (Rand- 

 wulst). The germ- ring also receives abun- 

 dant additions from the overlying primary 

 germ-layer, even at considerable distance 

 from the periphery. (See Reinhard, Arch, 

 f. Mikr. Anat., 1S98.) 



8. The whole germ-ring, extra-embryonic 

 as well as embryonic, contains both ento- 

 dermal and mesodermal elements (contra 

 H. V. Wilson and Samassa). 



9. Kupffer's vesicle arises, as in the 

 Salmonidfe, as a cavity completely shut in 

 by cells from the first. It is at first much 

 compressed horizontally and distinctly bi- 

 lobed. In embryos with a short tail it is still 

 to be seen near tip of the latter, strongly 

 suggesting neurenteric canal of Selachii. 

 A second vesicle, situated in yolk under the 

 posterior end of the embryo, appears slightly 

 in advance of Kupfier's vesicle and reaches 

 a size exceeding the latter. It is bounded 

 by periblast and perhaps contains more fluid 

 yolk for service of the growing end of em- 

 bryo. 



Respiratory and Breeding Habits of Polypterus 



Bichir. N. R. Harrixgton. 



On physiological grounds Polypterus is as 

 fully qualified for a ' lung-fish ' as are any 

 of the Dipnoans ; it has also striking resem- 

 blances in its circulatory and respiratory 

 system to the Urodela. These points were 

 demonstrated by means of mounted prepara- 

 tions, the injecting of which had been done 

 in the field principally by Dr. Reid Hunt. 



Beside the blood-supply to the lungs 

 (which is from the last branchial arch), the 

 dissections showed the very large glottis, or 

 ductus pneumaticus, by which the lungs 

 open ventral ly from the oesophagus. Un- 

 like the swimming-bladder of fishes in an- 

 other respect, both the lungs are entirely 

 invested with peritoneum, although one 

 of them, the right, does occupy the normal 

 position for an air-bladder, viz., between 

 the aorta and kidneys, on the one hand, and 

 the alimentary canal, on the other. The 

 mesenterj', however, iu which the left lobe 

 should be suspended, has almost entirely 

 degenerated, and this somewhat smaller lobe 

 lies entirely free in the body-cavity. 



It was pointed out that, while the strong- 

 est disproof of the Dipnoan ancestry of the 

 Amphibia lies in the paleontological evi- 

 dence which indicates that they are a par- 

 allel line, the same conclusion may be in- 



