2o8 BOTANICAL GAZETTE [march 



siderable part of each chromosome persists as an irregular compact 

 mass (fig. 85) . There is thus in these figures an indication that the 

 individuality of the chromosome is preserved even in the chromatin 

 network. 



GENERAL CONSIDERATIONS ON REDUCTION 



The important facts brought out in the present investigation con- 

 firm a number of conclusions put forward by the writer and others 

 during the past ten years, most of which have been the subject of 

 continual controversy. In a science like cytology so much depends 

 on the manipulation of the material and the interpretation of the 

 figures, that one need not be surprised at the diversity of views held 

 in respect to all the more important cytological problems. In the 

 present paper, by leaving out certain figures in the series, one can 

 produce several of the "reduction processes" heretofore published. 



The writer appears to have been the first to present a definite series 

 of observations to show that the first division after pseudo-reduction 

 is the real reduction division. A few previous reports had been pub- 

 lished, which were, however, largely guesses or assertions, w 

 definite evidence and sometimes even without drawings. 



In 1897, the writer presented his views on the reduction division 

 in the ovules of Lilium philadelphicum, advancing the definite con 

 elusions that the spirem is continuous and contains a single row of 

 chromatin granules which later undergo transverse fission; that the 

 continuous spirem doubles up and twists into twelve loops, the 

 reduction number, which then break apart at the inner ends opposite the 

 heads of the loops to form the twelve chromosomes; that during 

 metakinesis the two limbs of the chromosomes are pulled apart, fina J 



ithout 



• • 



and 



fir 



chromatin. In that paper figs. 1, 2, 2a, 4, 4a, 8, 8b, U, i lb > I2 > £ 

 21a, 22, 23, 23b, 34 , 35 formed a series for which only one interprea- 

 tion was possible. Only by leaving out fig. 4 could another interp ^ 

 tation be given, in which case the double spirem appearing later nug 

 be considered as conjugating instead of dividing. . 



In 1901 practically the same results were obtained for Erythrom 

 and in 1906 for the microsporocytes of Lilium tigrinum. 



