inherent sterilities the entire variety acts as one plant. Thus in the 

 variety Gertrude all plants true in vegetative descent from the original 

 seedling have sterile anthers quite as that seedling evidently had. A 

 variety which sets fruit freely to its own pollen in one locality will do 

 so elsewhere provided of course that incidental conditions do not inter- 

 fere. It may, however, be found that some varieties are somewhat 

 variable in their maleness or fe;maleness but as a rule rather uniform 

 behavior in inherent sterility is to be expected of clonal varieties. 



3Ietlio(ls of Study 



Hand jwllinations To determine the ability of a variety to produce 

 fruit and seed, hand pollinations were made using pollen known to be 

 capable of functioning in seed production. The most proper time for 

 pollination seems to be when the stigma is bent away from the crests 

 and the stigmatic lobe is somewhat drooping. This is usually from the 

 afternoon of the first day a flower is open to about the morning of the 

 third day when the flower parts begin to wilt. In malting pollinations 

 fresh anthers which were shedding pollen profusely, or as fully as is pos- 

 sible for them, were brushed over the stigmatic surfaces of all three 

 pistils. The flower was then tagged and a record made of the later 

 condition or development. 



In respect to the ability to produce fruit and seeds when properly 

 pollinated with good pollen varieties may be classed as follows: — 



A. Highly fruitful. 



Will produce fruit to good and viable pollen ("a" and "b" pol- 

 len as discussed later), with usually a high percentage of suc- 

 cesses. 



B. Feebly fruitful. 



Will produce fruit hut u.'-ually only to "a" pollen and with a 

 low percentage of successes. 



C. Entirely fruitless. 



No seeds will evidently be obtained to crosses made with any 

 pollen. 



The hand pollinations also constitute a test of the ability of pollen 

 to function when used on varieties known to be able to produce fruit. 

 The work was done in the display beds at The New York Botanical Gar- 

 den where it was impracticable to "bag" the flowers and hence stray 

 open pollinations could not be avoided. Under such conditions the pol- 

 lination tests are less conclusive for a judgment of pollen than for a 

 judgment of the ability to set fruit. To obtain further evidence regard- 

 ing the character and quality of pollen, laboratoiy studies were made. 



frennidnation of pollen. A direct study of the viability of pollen was 

 made by laboratory tests for germination. Preliminary tests showed 

 that good pollen of the irises germinates well on a medium made up of 

 1 gram agar agar and 15 grams cane sugar dissolved in 100 cc distilled 

 water. With 10 grams cane sugar good germination is also to be had. 



The method of making the germination tests is as follows: A large 

 drop of the melted medium is placed on a glass slide and allowed to 

 become cold and to solidify. Samples from the pollen to be tested are 

 then scattered or "planted" on the drop. The slide is numbered and 

 placed in a moist chamber and kept at room temperature for about 18 

 hours. The pollen is then examined with the aid of a microscope which 



