188 PROCEEDINGS OF THE ACADEMY OF [1897. 



length, slightly expanding to a lamina at the apex. The united por- 

 tions of the cyme-branch and petiole can be readily traced down- 

 ward towards the lower node. The separation of the edges results 

 in the usual square stem at this point. It may be remarked here 

 that one of the distinctions between Cornacese and Urnbelliferse is 

 the absence of an involucre. But here we find the tendency to dis- 

 pute even this character with the Umbelliferous order. The cases 

 sometimes given as illustrating such encroachment are not valid, for 

 the so-called involucre of Coruus florida results merely from a sec- 

 ond growth of bud scales. This occurs in many families of plants, 

 as, for instance, Pavia, Carya, Fraxihus and other genera, especially 

 in some hickories, where the rejuvenated bud scales are sometimes 

 as highly colored as in the Dogwood, and go by the name of " Hick- 

 ory-lilies." 



The appressed hair of the leaves and young branches of Cornus 

 stolonijera have been referred to as being " straight but fastened in 

 the middle, and thus appear appressed." But many will be noted 

 as having but one arm, which is as much appressed as if it had been 

 fastened at the middle. It is evidently a branched hair. The same 

 general law of sudden arrestation has evidently extended to the hair. 

 The apical growth, suddenly arrested, has driven the energy later- 

 ally, resulting in a horizontal growth on either side, as in the forma- 

 tion of the inflorescence. 



The tendency to abortion in many of the primordial parts, well 

 noted in Cornacece, may be illustrated here by some further facts 

 not hitherto specially noted. Though but one seed is produced, 

 two cells are seen in the transverse section of the carpel, and a 

 third has been occasionally observed. With a flower of four sepals, 

 four petals and four stamens, we may reasonably assume the early 

 existence of four carpels. But the stigma is usually described as 

 simple. If, however, we examine it just as the flower is fading, we 

 may note that it is really four-cleft, indicating certainly the four- 

 carpellary character of the ovarium. 



Another illustration of the prevalence of abortion is in the small 

 number of fruits which come to perfection in comparison with those 

 which fail. The flowers seem all perfect. It is safe to say that in 

 regard to seed-bearing nearly every flower is potential in all its 

 parts. Pollen and stigma are fully functionable. Indeed there is 

 scarcely a flower but is fertilized and seems to make a fair start to- 

 ward maturity. But the majority fall by the way. Though the 



