MATURATION MITOSIS IN AMPHIP.IA. 265 



been regarded as a second longitudinal split but what is really 

 the space between the two univalent chromosomes of a pair. 



The term " reduction division " was introduced by Weismann, 

 to denote a division that separates entire chromosomes, in con- 

 tradiction to "equation division" which halves longitudinally 

 each chromosome. Weismann, in his splendid conclusion that 

 such a division would be found to occur, unfortunately assumed 

 that there must be a doubling of the normal number of chromo- 

 somes before the maturation. This doubling does not occur, as 

 Boveri and Brauer were the first to demonstrate, but instead 

 there is in the synapsis stage, as I first showed, a union of chro- 

 mosomes end to end in pairs. This union is not due, as Riickert 

 suggested, to the chromatin spirem breaking into half the so- 

 matic number of chromosomes, since there is no continuous chro- 

 iiiatin spirem in the prophases of the maturation mitosis in any 

 of the objects studied by me, but is a union of chromosomes 

 that were disunited before. 



That a separation of entire univalent chromosomes in one of 

 the maturation divisions does occur, is shown by the following 

 studies : Riickert and Hacker for the ovogenesis of Copepoda, 

 finding the separation of entire univalent chromosomes to occur 

 in the second division ; while Lerat does not decide whether such 

 a division occurs, but points out that it is \h& first division which 

 is heterotypical. In the spermatogenesis of Gryllotalpa, by Vom 

 Rath ; of Hemiptera, by Henking, myself and Paulmier ; of Or- 

 thoptera by McClung, Sutton.and Wilcox. The first maturation 

 in the ovogenesis of Ophryotrocha, according to Korschelt's ac- 

 count, also separates entire univalent chromosomes, and very 

 clearly the first maturation in the spermatogenesis of Peripatus 

 according to my own observations. For the ovogenesis of Lim- 

 ncea this is maintained by Linville, but the chromosomes in this 

 object are unfavorable for decisive interpretation. In Isopoda 

 (Oniscus) the first division was shown to be reductional by Miss 

 Nichols. For TJialassciua and Zirpluea the same conclusions 

 were reached by Griffin ; and Lillie considers it possible that 

 Griffin's mode of interpretation may be applied to the egg of 

 Uuio. Van Winiwartc-r in a very excellent detailed study of these 

 stages in the ovogenesis of mammals (tfjiiio, Lcpns}, considers it 



