322 



NATURAL CROSS-FERtlLI^ATION IN INDIA. 



capsules in the first case contained more seed than those produced 

 by selfing, while those in the second case were much better. Selfed 

 seed gave fewer plants than free-flowering seed. No loss of vig- 

 our from continual selfing seems to have been noticed by this 

 investigator. Forms which stood next to one another and were 

 allowed to flower freely always crossed. Our observations at Piisa 

 in 1909 amply bear out Fruwirth's work on the frequent occurrence 

 of natural crossing in the opium of India. 



In 1909, 127 single plants of opium of many different types 

 were sown singly and the progeny in 1910 was examined and gave 

 rise to mixed offspring, only 9 plants breeding true. Splitting 

 took place in many directions, for example, into divided and nearly 

 entire leaves, round or oblong capsules, hairy and smooth stems, 

 in colour of the margin and eye of the petals and in the amount 

 of division of the petals. 



As regards the amount of division of the leaves two good cases 

 of splitting were observed : — 



No. 



J/136 

 4/134 



Parent plant, U)03. 



Prosenv, liilO. 



Leaves slightly diviiied 



Deeply divided. 

 29 



S'ightly divided 



43 

 Gl 



Usually smooth stemmed plants bred true. Almost all the 

 hairy stemmed plants gave smooth and variously hairy plants. 



A great deal of splitting was observed as regards the colour 

 of the petals (Plates X &XI). Out of 35 single plants with white 

 petals with an entire margin, 14 bred true and 21 split. Splitting 

 took place both with regard to colour and to the character of the 

 margin whether entire or divided. As a rule the flowers with entire 

 white petals which split gave white and light red petals. Plants 

 with divided petals generally bred true but colour differences often 

 came in. Thus out of 12 plants with white divided petals in 1909 

 only three bred true in every respect, but, as regards divided 

 petals only, 10 came true. 



