other groups appear to be less suceeessful. The Bvansias or crested 

 irises as a group produced fewer pods than did the apogons even to self- 

 ing and to crossing inside the group. The bearded irises only occasion- 

 ally set seed to their own pol'en. In crosses the Early I^warf and the 

 Tall Bearded varieties were more fertile both as seed and pollen pa- 

 rents than were the Intermediates which are mostly classed as "C". 



It is to be emphasized that the above report is a general survey 

 based on one season of work. In most cases the number of pollinations 

 possible to make with the material available is necessarily too low for 

 the results to be considered as final and conclusive. A larger numbei' 

 of pollinations and especially if fully controlled by bagging and if made 

 over a series of years would eliminate quite fully the various incidental 

 conditions which produce fruitlessness and would undoubtedly change 

 some of the pollination relations which were here all failures to success- 

 es. A further and .more comprehensive study of the germination of pol- 

 len would no doubt change somewhat the judgment of pollen for certain 

 varieties here classed as having poor pollen. 



In comparing evidence on the setting of fruit by irises obtained by 

 various persons differences in results are to be expected at least for 

 certain varieties. In Miss Sturtevant's published report (Bulletin No. 

 2, American Iris Society) the varieties Amas and Sarah had not set seed. 

 In the work here reported each of these varieties yielded a capsule con- 

 taining good seed to cross-pollination with \Vm. Marshall but tiieir gen- 

 eral behavior was such that they were classed as "feebly fruitful". Also 

 several varieties (Etta, Helge, Ingeborg and Spectabilis) classed by Miss 

 Sturtevant as "sterile" were in the tests at The New York Botanical 

 Garden found to be "feebly fruitful". Some varieties which were entire- 

 ly fruitless in the tests at the Botanical Garden were found to be "fer- 

 tile" by Miss Sturtevant (Arabes(!ue, Hector and White Knight are to be 

 mentioned." In all such comparisons positive results, if they involve 

 no errors in the naming of the clonal variety concerned, are to hv given 

 preference as indicating what the variety may be able to do. 



The survey indicates clearly that in general the clonal varieties of 

 irises fall into the four classes as designated by Miss Sturtevant. 

 namely: (1) varieties highly fertile both as seed and pollen parents, (2) 

 varieties that have only fertile pollen, (3) varieties that are pollen 

 sterile but able to set seed and (1) varieties that are sterile both in 

 pollen and in ability to set seed. Further studies combined with data 

 which growers and breeders of irises can furnish will in tiie course of 

 time more definitely determine the ability or inability of the different 

 varieties to function as seed or as pollen parents. It is clear that in the 

 cultivated irises sterility involving impotence of pollen or pistils, or of 

 both, is frequent and greatly limits breeding from seed. 



38 



