62 



The general subject is illustrated on the ])reee(ling page. In the 

 n])i)er figure we see the u])].)er and knver ])arts of the ovary of 

 Short ia (jalacifoHa {Diapciisiacciic) and Fcijoa Scllotviana (Myr- 

 hicrac). In each case the ni)i)er ])art shows ])arietal, the lower 

 ])art axile ])lacentati(.)n. This is analogous to the situation in 

 s\ ini)etal\', where the separate ti]is of the corolla represent the 

 primitive condition. The middle figure illustrates Dicfaiiiini.s' 

 alhiis and Ceiiinidcnia florihundd. We see the carjx'ls are more 

 nearly separate in the hud, more or less fused in the adult. In the 

 lower figure the develo])ment of i)lacentatiou in Sfcllarin media is 

 shown, essentiall_\' ])arietal at first, gradualU' fusing to central 

 placentation. h'inalK', after fertilization, tlie u])])er ])art of the 

 axis disappears. 



In October. 1923, in a Ih-ooklyn Ijotanie ( harden Lcajlcl, I 

 slated, " With the carpels closed to form a com])ound ])istil, the 

 ordinary axile ])lacentation results. Parietal placentation as in 

 rockrose, vi()let aud lH;)])py, and central placentation as in pink and 

 priuirose, are probably deri\-ed forms," again in October, 1925, 

 " The probable direction of evolution in fiowering ])lants may be 

 briefly summarized: axile i)lacentation to parietal or central." 

 Some time later a discussion with Dr. 1'. A. Kydberg caused me to 

 doubt the above conclusions aud to begin a stud\- of the subject 

 of ])laeentation. (draduall}- it bi'cauie clear that i)revailing views 

 had been mistaken, and also that a])plication of jjlacentation to the 

 classification of dicotyledons in various res])ects siui])lifies the 

 s}'steni. Hutchinson stated iu 1926, " The parietal type of ])lacen- 

 tation generall\- i)rece(le(l the axile, basal or ai)ical ty])es, the axile 

 being the iinal aud uiost etficieut condition." I>ut little a])i)lication 

 of this principle a])p(.'ars to have been made in Hutchinson's book 

 on Dicotyledons, as when he deri\es J.yllirulrs, Ccraiiialcs, and 

 (,'riiliiiiiiilcs from Caryopliylhdcs. The CaryoplwUaccac a])i)ear to 

 have had a better ])osition in the old systems, namely near I-nnik- 

 CiiMccac. In the 1 lutchinson swstem, howex'er, numerous groups 

 with ])arietal i)laceutation just follow MiUjiiolialvs. With these 

 groui)S might well be joined Rhocadalcs, riohdcs. and other 

 orders from the herbaceous side of his svstem of classification; and 

 similarly in other cases, by uniting along \arious lines Ilulcbin- 

 sou's gr(.)U])S " arboi-escent " aud " herl)ace(.)us," we believe a more 



