NO. 1431. BREEDING HABITS AND EGG OF PIPEFISH— GUDG Eli. 481 



teleostean with the fourth (pole-av^oiding, meridional) furrow of the 

 frog-. For this interpretation of Rauber see Wilson (1891, pp. 211-i^l5). 

 Agassiz and Whitman (1S85) think that the amphibian equatorial 

 furrow has l)ecome vertical in the Teleost, and that the horizontal 

 division of the four central cells of the sixteen-celled stage into four 

 outer and four inner lying cells is the tirst equatorial segmentation. 

 With this latter statement Kopsch (1901), from his work on Brloiic, is 

 in full accorel. Brook (18ST) describes, from sections of herring- 

 eggs (Plate XIII, tig. 9), ati e({uatorial segmentation separating the 

 four blastomeres from the pei-il)last. List (1887, Plate XXXI, tigs. 1 

 and 5) tinds the second furrow in (^i-rn/fohrux to be equatorial, and 

 says that Kupfi'er found the same in the herring. In Crist Ice j>><^ 

 Fusari (1890, tigs. 1 and 5, Plates I and III) hnds that in the sixteen- 

 celled stage, all the cells are united at the base, but the next division 

 sets sixteen central cells free from the yolk and from sixteen periph- 

 eral cells. This he calls the equatorial division. Wilson (1891, p. 

 215) agrees with Rauber (see above). Samassa (1896), in the segmen- 

 tation of Salmonoids, tinds as a rule that an equatorial division follows 

 the eight-celled stage, although it sometuues comes earlier. 



a h e a r f 



Fii;. 2. — Kggs of the Tritons ix thk eight-cei.led stage. (After Groxross.) 



An equatorial segmentation has been pointed out in certain eight- 

 celled blastoderms of S!j>ho.stoiiia^ and this gives them a very decided 

 resemblance to the upper surface of dividing amphibian eggs. 

 Gronross (1890) (see Hertwig's Handbuch) gives a series of hgures 

 for the eggs of Tritons to which the tigures above noted show very 

 striking resemblances. The Tritons have eggs with relatively large 

 amounts of yolk and in them the segmentation approaches the mero- 

 blastic condition. The text figure reproduces some of the more strik- 

 ing forms to which reference will be made. The resemblance is so 

 striking that no extended comparison is called for. With Gronross's 

 fig. a., compare fig. 10, Plate V, and also Henneguy's (1888) fig. 39. 

 They are almost identical. For a figure which almost duplicates his 

 figs, h and c, see fig. 13 on the next plate. Among drawings not 

 included in the plates is one almost identical with his fig. d. Again, 

 figs, e and f are very similar to fig. 9. The comparison might be 

 extended further, l)ut this is sutficient to show the very striking simi- 

 larity between these two forms. That we have here an analogous 

 segmentation is beyond question. The segmentation in the pipefish 



