May, 1905.] The Nature of the Reduction Division. 337 



gametophytes with a true alternation of generations tlie}^ are 

 retained in the ovary. The plant life cycles together with a 

 typical animal are compared in Plate XXI. 



Among the interesting things which have recently come to 

 light and which appear to have their basis in the phenomena of 

 the reduction division is Mendel's law of heredity in hybrids. 

 The operation of this law can be explained on the hypothesis of 

 pure sex cells. In 1897 the writer worked out in detail the 

 reduction division which takes place in the ovule of Lilium 

 philadelphicum. Although the development and subsequent 

 division of the chromosome was followed out in considerable 

 detail in this work, the facts presented were not accepted by a 

 number of botanists, as admitted by Strasburger, because of 

 supposed authority in the opposite direction. So soon, however, 

 as Mendel's law was rediscovered it became self-evident that 

 belief in a qualitative or true reduction division of some kind was 

 necessary if the whole chromatin hypothesis was not to fall to 

 "the ground. Accordingly a re-investigation by some of the fore- 

 most cytologists, among them Strasburger and Farmer, resulted 

 in a confirmation and acceptance of the propositions presented 

 in my papers on Lilium and Erythronium, as also of similar work 

 done previously by a number of zoologists. 



Gsriis ensE 



h 



Fig. 3. Diagramatic representation of the transverse division of a 

 chromosome. 



If then a transverse division of the chromosomes occurs dur- 

 ing the reduction karyokinesis each of the chromosomes resulting 

 from the process of pseudo-reduction may be regarded theoret- 

 ically to be made up of a pair of chromosomes, one being a male 

 chromosome and the other a female chromosome, joined end to 

 end. There are also some obrsevations which appear to indicate 

 that this actually takes place. In such cases then as in Lilium 

 and Erythronium the formation and nature of the twisted loop 

 can be diagramatically represented as in Fig. 3, a-e. In the fol- 

 lowing division in which a longitudinal splitting occurs the 

 daughter halves of the chromosomes formed early in the previous 

 division become separated (Fig. 3, f-h). 



Mendel's law of heredity, so far as it has a direct bearii;g on 

 the nature of the reduction division, may be briefly stated as fol- 

 lows: When two organisms differing in some character are 



