Studies of Teratological Phenomena. 



109 



casual observation can determine, also shows this slight variability, and 

 in practically the same degree. In an omitted table, the frequency 

 distribution of these variations of parts per flower and the number of 

 the abnormal flowers to each plant were calculated. Out of 11 •475 

 flowers collected from 459 normal segregate plants or their progeny, 

 there were 11*079 normal to 396 abnormal, or over 3*5 per cent of 

 the latter. These abnormal flowers, representing 240 of the 459 plants 



Fig. 20. Young inflorescences of 304 — 1, 304 X 402 and 402. 



herbarium specimens. 



Photographed from 



concerned, had 608 extra floral parts (sepals, petals, etc.) or about 

 1*5 part per flower. Their absence from the remaining 219 plants is 

 of no significance, as each plant of the whole 459 was represented by 

 only 25 flowers, and as abnormal ones are more common on branches 

 just entering the blooming period, I might easily have overlooked them 

 on some plants. Plants on which two or three such flowers had been 

 found were selfed and their progeny grown, and from an examination 

 of these, I concluded this variation was not heritable (Fig. 21). 



As to whether the flowers of the aa segregates, in respect to 

 number of parts, are as normal as those of the 402 grandparent, is an 



