150 BROOKLYN BOTANIC GARDEN MEMOIRS 



Thus plant 2, used as a female, was fertile with pollen from plants 4, 

 18, 41, 44 and 52, and sterile with plants 9, 22 and 23; but pollen 

 from plant 2 was fertile on plants i, 3, 4, 5, 7, ii, 18, 20, 28 and 29, 

 and sterile on plants 9, 10, 13, 25 and 27. It is clear, therefore, th^t 

 instead of the 8 matings on plant 2 that Table I appears to show, 

 there are really 21, the 3 reciprocals of course being counted but once. 



These tables were combined for analysis. In the interest of 

 economy of space only one is shown, however, since the second can 

 easily be made from the first. 



The four exceptions in this huge set of matings are in reality 

 negligible. Matings 15 X 44 and 31 X 36 were sterile, though they 

 do not belong to the same class. Plant 15 was sterile to 4 plants of 

 Class A and fertile to 2 plants of Class B, 3 plants of Class C, and to 

 the isolated individuals forming classes D and F. It is unquestionably 

 a member of Class A. Plant 44 was sterile to 7 individuals in Class C 

 and fertile to 17 plants of Class A, 12 plants of Class B and to the 

 singletons forming classes D, E and F. This evidence places it un- 

 mistakably as a member of Class C. Plant 31 is also a member of 

 Class C as evidenced by 3 sterile matings within that class and by 

 fertile matings with i plant of Class A and 3 plants of Class B. Plant 

 36 is like plant 15 thrown into Class A by its sterility with 3 others of 

 that class, and by its fertility with 3 individuals of Class B, with 2 of 

 Class C, and with the lone plant of Class D. In view of this evidence 

 and the fact that in these two matings but one pollination was made 

 in each case, they are much more likely to be errors of record or of 

 technique than true exceptions to our classification. 



The other two exceptions, matings 45 X 18 and 33 X 46, were 

 fertile where from the evidence of numerous other matings they should 

 have been sterile. Here again but one pollination was made in each 

 case; and, coincidence though it may be, each pollination was the last 

 mating made on that particular plant. What is more probable than 

 that this is a pseudo-fertility appearing during the wane of the flower- 

 ing season of the two mother plants, No. 45 and No. 33? 



Six groups appear in Table II, but there is proof of the existence 

 of only five. Groups A, B, C, D and E are definitely established. 

 Plant II, on the other hand, is an isolated individual rather than a 

 class. It does not belong to groups A, B or C; but unfortunately it 

 was not crossed either with Class D (plant 20) or with Class E (plant 

 43), hence one cannot say that it does not fall into one or the other of 

 these two classes. 



In the three large groups the distribution of individuals is 22, 16 

 and 12. About all that can be said about the type of this distdbjiUion 

 is that the classes are not of equal size. On the other hand, it is 



