THE DOUBLE STOCK, ITS HISTORY AND BEHAVIOUR. 463 



(Mentioned by Decaisne and Naudin * as a practice among French 

 gardeners.) Chate states that he drew his conclusion from the 

 results he obtained on saving the seeds from the upper and lower 

 ends of the pods separately. I have repeated Chate 's experiments, 

 but I have not found that the results give any confirmation of his 

 view. The proportion of doubles obtained from the two halves 

 of the pods was found, on the average, to be the same, an excess 

 of singles sometimes obtained from seeds taken from the lower end 

 being counterbalanced another time by an excess among plants which 

 had been grown from seed taken from the top end.f The greater 

 number of doubles obtained by Chate from the lower ends of the 

 pods was no doubt accidental, and would not have held good if the 

 experiment had been carried out on a large enough scale. 



The same explanation probably accounts for the view quoted by 

 Naudin and Decaisne, if indeed it had any foundation in fact. In 

 this connexion I may remark that in the Stock it is not unusual to 

 find what appears to be a very irregular distribution of the doubHng 

 factors among the ovules of individual pods ; hence a large sowing 

 is often essential in order to determine whether the result obtained 

 from a sample sowing represents the true ratio, or an accidental 

 departure from that ratio. 



10. Selection for sowing of seed harvested from a summer-flowering 

 Stock which has survived the winter (mentioned by Oberdieck).J 

 This view again presupposes that the double character is one capable 

 of being produced as a direct result of a change in environmental 

 conditions, and therefore need not be further discussed. 



11. Selection of those seedlings which arise from energetically 

 germinating seed [e.g. seed germinating in three to four days), and the 

 discarding of those of the same sorts which germinate slowly {e.g, 

 seed taking nine to ten days to germinate) (advocated by Nobbe,§ 

 1887). I am unable to decide from Nobbe's account whether he holds 

 that the rate of germination in itself hSiS an effect on the double character, 

 of such a nature, that in some cases a seed germinating very slowly 

 produces a single, which, if it had germinated more quickly, would 

 have given rise to a double ; or whether his statements are to be 

 interpreted as meaning that the embryos of seeds destined to produce 

 doubles are capable of more rapid germination than those of seeds 

 which will give rise to singles. The first -mentioned supposition falls 

 into the same category as the views given under 7, 8, and 10, and may 

 be dismissed for the same reasons. As regards the second possibiHty — 

 that the germination of seeds which will in any case produce doubles, 

 as compared with seed which will give singles, is more rapid — I am 

 disposed to think it unhkely, and that in any case Nobbe's observation 

 is not on a sufficiently large scale to be conclusive. To obtain perfectly 



* Manuel de V Amateur des Jaydins, Paris, 1862. 



t See Saunders, " The Breeding of Double Flowers," IVe Conference Inter- 

 nationale de Genetique [Comptes rendus et Rapports), Paris, 191 1. 

 X loc. cit. 



§ Bot. Centralblatt, Bd, 32,^1887, p. 253. 



