DOUBLE FLOWERS. 



475 



made in the case of the Chinese Primrose (Primula sinensis) and the 

 Wallflower [Cheiranthus Cheiri) the first crossbred generation consists 

 of singles.* 



Tho'Ugh we can learn something of the relation of double to single 

 by observation of the crossbreds obtained by mating the two forms 

 together, it is only by the statistical examination of the later genera- 

 tions derived from these first crossbreds that we are able fully tO' 

 understand the conditions underlying the appearance of doubles in 

 each case. Breeding investigations of this kind show beyond all doubt 

 that doubling is a character which in many cases is transmitted from 

 parent to offspring in accordance with definite principles, or laws of 

 heredity, and that these principles are inherent in the nature of the 

 organism itself and are independent of external conditions. As 

 showing the nature of the evidence upon which this statement is based 

 I may briefly state the facts in the case of certain types which have 

 been studied in some detail. 



If the single plants produced by crossing a single with the present- 

 day type of double in Primula sinensis are self- or inter-bred they are 

 found to yield in the next generation a mixture consisting on the 

 average of three single plants to one double, f The doubles so obtained, 

 and their posterity, breed true to doubleness. A certain proportion 

 of their single sisters will be found to be incapable of throwing 

 doubles, but on the average two of every three of these, singles will be 

 capable of yielding either 25 or 50 per cent, of doubles according as 

 the pollination is made with the pollen of a single or a double, and 

 such doubles also breed true. In fact, a double once obtained, how- 

 ever bred, henceforth breeds true, if protected from the pollen of 

 singles. The behaviour of the single plants, on the other hand, 

 cannot be predicted from inspection. Their behaviour varies according 

 to the way in which they have been bred, although in outward appear- 

 ance they are all similar. Only by observing the offspring can we 

 discover whether a given single is pure-bred or crossbred in constitu- 

 tion, for, if pure-bred it will yield only singles, if crossbred, it will 

 give a definite percentage of doubles (25 per cent, if self -fertilized, 

 50 per cent, if crossed with a double). In the double of the Primula 

 we have an illustration of a very simple case of Mendelian inheritance, 

 the essential feature of which is that in the case of two opposite 

 characteristics, one due to the presence and the other to the absence 

 of a certain factor or combination oi factors, the reproductive cells, 

 likewise, either carry or lack each of the factors concerned. Hence 

 ^ when, as here, a particular characteristic appears to be determined by 

 i one factor only, a plant may be derived from any one of the following 

 I combinations : 



* Exceptionally, however, an crossbred in the Wallflower may show an 

 extra structure in a flower here and there in the form of a slender filament 

 behind the petal, due to splitting (fig. 171, g). 



I t See Gregory, "Experiments with Primula sinensis," Journal of Genetics, 



1 vol. i. No. 2. 



