270 PROCEEDINGS OF THE ACADEMY OF [1890. 



extend to distinctively American species, though Old-World species 

 retained the peculiarities when translated here. I now find in 

 Pyonanthemum similar conditions. P. lanceolatum Pursh, a large 

 clump of which from one original plant growing in my garden pro- 

 duces flowers apparently ])erfect, wholly fails to perfect seed. 

 Close beside it is a clump of P. wuticum Fers. with no trace of stamens, 

 hut with an exserted and in every way perfect pistil, seeding 

 abundantly. On close examination it is evident that the pistil in 

 the former is not as i)erfect as it appears to be. It is never exserted, 

 nor do the stigmatic lobes perfectly expand. There is no doubt that 

 these plants practically represent staminate and pistillate i)lants 

 respectively. 



The abundant seeding of P. muticwn under the above detailed 

 circumstances, could only result from the abundant pollen of P. 

 lanceolatum, and the seeds produced by the former must of necessity 

 produce hybrid plants. 



Prof. Gray notices in Synoptical Flora a number of variations near 

 P. muticum, and of one, P. leptodon, remarks " perhaps a hybrid be- 

 between P. muticum var. pilosum, and P. Tullia var. dubium." The 

 observations here recorded add greater probability to the suggestion 

 of the hybrid origin of intermediate forms in this genus, by pointing 

 out the exact method under which hybridism is brought about. 



It may be repeated here that, as noted by the author long since, 

 cross-fertilization must be regarded as the hand-maid of heredity, 

 rather than as a factor in the evolution of form. Its duty is to 

 bring nearer home that which has wandered. 



Self-Fertilizing Flowers. 



A large number of flowers are so arranged as to render self-ferti- 

 lization difficult, and in many cases impossible. Numerous observ- 

 ers have placed on i-ecoi'd what they have discovered in the line of 

 these facts, and have rendered inestimable service to science. There 

 may be a question whether the interpretation of such facts be legit- 

 imate ; there can be no question as to the value of the facts them- 

 selves. 



Observations on the other side have not been as numerous, yet 

 they seem no less interesting. The number that are not only adapted 

 to self-fertilization, but actually self-fertilized, seems as large as that 

 of the other class. We may leave the interpretation till the record 

 is more complete. 



