Double Flowers and Flowerheads. 195 



(Cineraria), the transformation of the httle yellow disc- 

 florets into long white tubes (Pyrethnmi, see Fig. 36) 

 etc. We shall however leave such cases out of considera- 

 tion ; they may be regarded provisionally as cases of 

 spurious doubling. 



The genuine 'Vloubling," on the other hand, as ex- 

 hibited by the most diverse species, presents a very 

 marked agreement with the conditions found in Cliry- 

 santheuiuin segetiun. Indications of a tendency to "doub- 

 ling" occur both in forms of which a double variety is 

 not offered by seedsmen and in those of which such 

 are already on the market. For instance in 1892 I ob- 

 served occasional tube-florets more or less completely 

 transformed into ligulate florets in a culture of Bideus 

 (jrandiflora in my garden. In other cases the variation 

 is only seen when curves are plotted. For example I 

 obtained the following very asymmetrical curve from 

 a culture of the single variety of Clirysaiifhenium coro- 

 nariimi, a favorite garden plant whose double form 

 has long been known (Fig. 35). 130 flowers on 25 

 plants of a single crop were recorded, the flower at the 

 top of the main stem and those on the primary branches 

 alone being taken into account. I found : 



Ligulate florets 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 

 Inflorescences 10 1 2 2 12 25 19 21 IS 14 6 7 2 3 



That is to say, 18 on one side of the mean and S7 on the 

 other, with a faint indication of a second maximum at 

 the next figure in the Braun-Sciiimper series, 21. It is 

 clear that the (l(ml)le variety of this species could prol)- 

 ably be obtained from these plants, exactly in the same 

 way as in C. segefiim. 



These considerations evidently lead to the hypothesis 

 that the secondary maxima on the positive side of Lud- 



