82 CROSSING. 



white in platyceras. In the second generation the variability 

 was stupendous. (Fig. 1 1). A. whole list of novel characters was 

 found. Some plants had purplish stems, a character unknown 

 in either parent-species. As to the flowers, a number of Fa 

 plants had petals of unequal size, three long ones alternating 

 with three smaller ones. At least two of the plants had only 

 three petals instead of six. New colours were many, including 

 salmon pink. Double flowers in diverse grades were found on 

 several plants, one plant having almost all the stamens con- 

 verted into petals. Several plants had more or less laciniated 

 petals, and in one plant the petals where as much fringed, as in 

 the laciniated domestic species related to Papaver somniferum. 

 Finally, polycephaly was noted in at least two plants, a trans- 

 formation of stamens into ovaries, containing ovules. 



This origin of these novelties in the second generation of a 

 species cross in Argemone shows how wholly similar aberrant 

 characters in other Papaveracea have probably originated. 

 Polycephalic poppies are known in several species. As we know 

 how many crosses horticulturists have made in those plants to 

 obtain variation, we need not assume "mutation" to explain 

 the origin of such novelties. 



It happens to be the custom of the horticulturists to regard 

 one of the species crossed as more important than the other, 

 in almost every instance where crossing is resorted to. From a 

 - genetic point of view this practice is on the whole defendable. 

 Species-hybrids are not as a rule the direct parents of new val- 

 uable horticultural novelties. The real process is generally as 

 follows. A promising species is taken into cultivation, and pos- 

 sible variation is carefully watched for. Every pcfesible cross is 

 tried, and if hybrids are obtained, these are crossed back into 

 the species. Variability is the result, and from now on novelties 

 begin to crop out. These are separated, and variously com- 

 bined. Newly imported related species are eagerly sought after, 

 as furnishing material for crosses. From personal observation 

 of the work of horticulturists, we know that the object of cross- 

 breeding with a newly imported species does not appear to be 



