264 THOS. II. MONTGOMERY. 



chromosomes ; he also failed to describe the true longitudinal 

 split in each arm of the chromosomes, though certain of his 

 figures show it indistinctly. McGregor confirmed for Amphiuma 

 Meves' account for Salamandra, but likewise failed to show how 

 the bivalent chromosomes are produced ; his Fig. 7, PI. IV., 

 shows the true longitudinal split, but it is not demonstrated by 

 his figures that this becomes the space enclosed by the definitive 

 chromosome. Eisen likewise interprets, for BdtracJwscps, the 

 heterotypic division as an equation division, though his "bouquet 

 stage" is essentially similar in its details to my Fig. 2, and though 

 his results might equally well be interpreted as speaking for 'a 

 transverse division. Kingsbury's account for Desmognathus is 

 clearly a careful study ; his Figs. 4 and 5 may be compared with 

 my Figs. 4 and 3 ; he found the polarity of the chromosomes 

 in the nucleus, saw the true longitudinal split (his Fig. 6), but 

 illustrates no stages to show that this split is connected with the 

 space in the definitive chromosomes. Janssen's account for Triton 

 appears to be the most detailed and careful, his Figs. 4, 5 and 

 32 show appearances of the early chromosomes just as I have 

 found ; he describes the synapsis stage where the formation of 

 bivalent chromosomes takes place ; he was the first to clearly 

 recognize this stage in the Amphibia, but holds the bivalent 

 chromosomes undergo two longitudinal splittings after their 

 transverse segmentation from a continuous chromatin spirem. 



In the ovogenesis of Amphibia, to mention only two more 

 recent studies, Lebrun concludes that both maturation divisions 

 are equational, though more on the basis of a lengthy discussion 

 of the definitive forms of the chromosomes than of a description 

 of the early formation stages. Miss King, for Bafo, was unable 

 to determine whether the heterotypic division is transverse or 

 equational. 



The whole question is one of a careful interpretation of the 

 early stages of the chromosomes, and the mode of formation of the 

 bivalent chromosomes. The workers on this subject have been 

 satisfied for the most part to know that in the spermatocytes the 

 chromosomes are bivalent, without describing the mode of union 

 of every two univalent chromosomes. In Dcsinognathns there is 

 one true longitudinal split, and besides that what has heretofore 



