Febeuaby 11, 1898.] 



SCIENCE. 



197 



its eyes and brain, etc., never formed out of 

 old tissue, but always from new tissues. 



In one case two heads formed on oppo- 

 site sides of a short cross-piece that had 

 been cut from the middle of the body. The 

 head, therefore, had the normal orientation 

 of the piece, while the other, turned in the 

 opposite direction, had its orientation ex- 

 actly reversed. 



Regeneration and Grafting in Cordylopliora. 



G. Lefbvee. (Presented by E. A. An- 

 drews.) 



The stems of Cordylophora, when cut in 

 pieces, exhibit the heteromorphic formation 

 of hydranths, as has already been observed 

 in this hydroid. 



The ccBuosarc regenerates the new hy- 

 dranth at the cut end by a distinct process 

 of budding, growing out beyond the old 

 perisarc into a knob-like projection which 

 acquires the rudiments of tentacles in from 

 36 to 48 hours. This is not merely a direct 

 transformation of the tissues of the stem 

 into the body portion of the hydranth un- 

 accompanied by growth, as has been de- 

 scribed for other Tubularian hydroids, but 

 the process is in truth a regeneration or new 

 formation. 



A piece of stem invariably regenerates a 

 new hydranth at each end, even when lying 

 on the bottom of a dish. Usually a foot is 

 formed when a stem is brought in contact 

 with a solid object, but in several cases it 

 was found that a hydranth arose at the end 

 which was firmly attached to the dish, the 

 hypostome acting as the organ of attach- 

 ment. This inverted hydranth did not at- 

 tain to complete development, but it was 

 a distinct hydranth provided with several 

 short tentacles. 



Only negative results were obtained from, 

 isolated tentacles, as no regeneration took 

 place, the tentacle soon contracting into a 

 rounded mass and dying. 



Grafting may be successfully performed 



on the stems of Cordylophora. When 

 freshly cut pieces are brought into contact, 

 end to end, a firm, complete, permanent 

 union takes place. Ectoderm unites with 

 ectoderm, endoderm with endoderm. There 

 is no polar differentiation in regard to the 

 ability of the stems to fuse with each other, 

 and in the experiments which were made, 

 series of fused pieces were obtained repre- 

 senting all the possible combinations of the 

 two poles. The united stems did not 

 eventually break apart, but remained intact 

 until they finally died in Mo after several 

 days. At the point of union between two 

 pieces a lateral branch was given off in 

 many cases, each portion apparently con- 

 tributing equally to the branch. 



A Recent Variety of the Flatfish, and its Bear- 

 ing upon the Question of Discontinuous Vari- 

 ation. H. 0. BuMPUS. 

 It was shown that within the past five or 

 six j'ears the lower side of the flatfishes 

 {Pleuronedes Amerieanus') from Woods Holl, 

 Narragansett Bay and Long Island Sound 

 has, with great frequency, become deeply 

 pigmented over more or less]|definite tracts. 

 The abrupt appearance of a large number 

 of individuals, varying in a definite direc- 

 tion, bears directly upon many current 

 speculations of organic evolution. It was 

 claimed that the variation being so wide- 

 spread must have been the result of some 

 environmental stimulus upon the germ, 

 since the arrangement of the color precludes 

 the possibility of its being the result of the 

 direct action of light, and there is evidence to 

 prove that the appearance of ' piebald' speci- 

 mens was not due to the invasion of piebald 

 fish from other localities. 



It was also claimed that the process of 

 natural selection could not have been mate- 

 rially instrumental, since the presence of 

 the piebald specimens was first indicated 

 by a large number of young fish, and be- 

 cause the time has been too brief for nat- 



