THE DOUBLE STOCK, ITS HISTORY AND BEHAVIOUR. 465 



support of Chat^'s view, it seems scarcely profitable to discuss it 

 further. 



13. Selection of seed from pods which are quite closed at the apex, 

 those being discarded which are open or only half closed at the upper 

 end (suggested by Papanck).* This idea is purely speculative ; as 

 the author naively explains, selection based on differences in the seed 

 characters is difficult to carry out, and he therefore directed his atten- 

 tion, not, like Professor Nobbe, to the seeds, but to the pods, since, if 

 differences in pod characters were found to be associated with singleness 

 and doubleness in the plants derived from them, this distinction would 

 afford a practicable method of selection. From the results of his first 

 experiment he was disposed to think that doubles were produced by 

 seeds from the closed pods, singles from those that were open, but adds 

 that he did not regard this first experiment as conclusive. Confirmation 

 of this suggestion was not obtained, as an accident prevented the 

 carrying through of a second test. A side point noticed by Papanck 

 was that seeds from closed pods appeared to germinate three or four 

 days later than those from open ones. This works out, on Papanck' s 

 view, that the doubles germinate more slowly than the singles, whereas 

 Nobbe came to the opposite conclusion. This view of Papanck's, 

 however, which is admittedly purely speculative and supported only 

 by very slight evidence which he himself hesitated to accept, I think 

 we may very well leave out of account, regarding it as merely one more 

 of the many unsuccessful attempts to find in some general outward and 

 visible sign an indication of a character already determined, but not 

 yet apparent in the embryo. 



Before leaving the subject of the possibility of identifying the future 

 doubles by the seed characters, I may add that so far as my experience 

 goes this is only feasible in one strain, viz. the sulphur- white, and only 

 in one of the races belonging to this strain. The sulphur-whites are of 

 special interest from the fact that they are not only ever-sporting as 

 regards singleness and doubleness, but also as regards flower colour ; 

 and further, on account of a certain coupling of the plastid colour 

 character with singleness and doubleness. The singles are white, both 

 sap and plastids being colourless ; a very few of the doubles are also 

 white, but almost all are cream, owing to the presence of cream-coloured 

 plastids in a colourless sap. This strain is found both with the small 

 lumpy seeds characteristic of the ordinary true-breeding white, and 

 with the large rounded flat seeds characteristic of the true-breeding 

 cream. In this latter case two shades of colour may be distinguished 

 in well-ripened seeds, one brighter and yellower, the other duller and 

 browner. On sorting the seeds into these two classes it was found that 

 the creams, and therefore nearly all the doubles, were produced by the 

 yellower seeds ; the whites, and therefore almost exclusively singles, 

 from the browner seeds. By this method of sorting it is therefore 

 possible in this case to obtain a fairly complete separation of the doubles 

 from the singles. The actual numbers recorded in the single experiment 

 * Wiener lUustrirte Garten-Zeitung, 1899, p. 389. 



