94 



Rediiplicatiofi Series in Siceet Peas 



be pointed out however that this scheme is not incompatible with the 

 n : 1 basis. The form of the series, whether (n- 1) : 1 or n : 1, might 

 depend upon whether the first division in the quadrant were a periclinal 

 or an anticlinal division. In the one case (Fig. 1) we should get the 

 (n - 1) : 1 : 1 : (m - 1) series, and in the second case (Fig. 2) the 



1st division 

 f 1st division 



2nd division 



Fig. 1. 



Fig. 2. 



n : 1 : 1 : n series. Further possible ratios also depend upon whether 

 the first division in the other two quadrants is periclinal or anticlinal. 

 Indeed it is obvious that there are numerous possibilities which may 

 perhaps repay discussion when more experimental data are available. 

 All that can be stated positively at present is that the cases hitherto 

 worked out in the sweet pea fit in with the hypothesis that the number 

 of cells in the reduplicated series is some power of 2 where only two 

 factors are concerned. But where three factors are concerned this is 

 certainly not true. The value of the primary reduplications is evi- 

 dently altered, and there would seem to be some process whereby these 

 reduplications react upon one another. Where so many points remain 



American variety with coloured aleurone and horny endosperm, and a Chinese variety with 

 white aleurone and waxy endosperm. By means of other statistics Collins is at pains to 

 prove that the reduplication phenomena in maize are of a highly irregular nature. Much 

 stress however cannot be laid upon these results as the author is evidently dealing with 

 dominant as well as recessive whites in his experiments though this point does not appear 

 to be specifically recognised by him. It is probable that a more careful genetic analysis 

 of the whites which he uses would help to clear up the apparent irregularities in his 

 results. 



