60 



THE AMERICAN BOTANIST 



shell; the other two carpels are each but one-half the size of 

 such. As to the internal structure, we have not investigated 

 further. 



The other walnut is made up of four distinct carpels all of 

 about equal size and every one with a more or less notable 

 false partition reaching inward and dividing the embryo coty- 

 ledons into eight more or less imperfect divisions. The embryo 

 itself, at the radical end, has the shape of a four-sided pyramid. 

 As already noted, we need not suppose that all abnormal fruits 

 are to be interpreted as reversions to primitive or atavistic 

 types. Years ago, we found a ''fruit" of a member of the same 

 famih', Hicoria alba, the common hickory nut, that had four 

 carpels or a double nut with a perfect partition between two 

 perfectly normal healthy seeds. This condition would possibly 

 have to be interpreted as a reversion to a still more primitive 

 type. It seems quite satisfactory to consider all these as just 

 abnormal conditions for which we need not rack our brains to 

 find an explanation with very much hope of profit. 



Xotre Dame, Ind. 



[In addition to what has been said above about abnormal 

 fruits in the Juglandaceae, we may add that we have in our 

 possession double fruits - of both the black walnut {Juglans 

 nigra) and the butternut (Jiiglaiis ciiicra). In these there are 

 two perfectly formed seeds enclosed in a single husk. \A> also 

 have a black walnut with three cotyledons. This latter speci- 

 men seems in no Avay verv dift'erent from seeds of other kinds 

 which occasionally produce seedlings with three cotvledons. 

 In certain cases De\^ries was able to breed a race of tricoty- 

 ledons from such beginnings. — Ed.] 



I 



