GABRIELLE L. C. HOWARD. 97 



In Type 51 the position of the greatest wddth has been 

 determined for a large number of leaves, and is always half-way 

 between the apex and the base. In Type 9 it is lower at a 

 point two-fifths from the base of the leaf. Determinations of 

 this point made in the F4 generation on cultures derived from 

 694 showed that all were uniform in this respect and resembled 

 Type 9 exactly, although in size and ratio length/breadth they 

 were very like Type 51. On Plate XX some of these leaves 

 are shown, contrasted with a leaf of Type 51. Apparently this 

 character can be inherited quite independently of the ratio 

 length/ breadth and it is possible to combine in the same leaf the 

 ratio length breadth of one parent with the greatest width in the 

 same position as in the other parent. Similarly, very distinct 

 segregation as regards this character was observed in the cross 

 Type 23 < Type 38. A large number of cultures have been 

 measured with reference to this point, but as the number of 

 tables is already great, the results will not be published until 

 the fuller investigation of the factors concerned with the 

 shape of the leaf is completed. The measurements were carried 

 out by moving a steel measure at right angles up the leaf until 

 it denoted the greatest width, when its position was marked by 

 inserting a needle into the mid-rib. The distance between this 

 point and the base of the leaf was then determined. 



The ratio length/breadth and the position of the greatest 

 width are sufftcient to determine the general form of the leaf in 

 all cases. If we further imagine pieces of varying size to 

 be cut out at the apex and base of the leaf we are able to 

 reproduce the leaves of all the existing types. If we postulate 

 the existence of independent factors the effect of each of which 

 is to cause a different indentation in the outline, all the facts 

 which have been discovered during this investigation can be 

 explained. 



The most striking example is that given by the cross 

 Type 23 x Type 38 (Table XXVIII). Both these forms 

 have sessile leaves, but the amount of indentation near 



