200 THE AMERICAN MONTHLY [July 



merman, Bouuier, Hicks, aud others. Some careful stud- 

 ies were made by Frank, Sempolvvski, Hoelmel, Schroeder, 

 Harz, Abraham, and D'Arbaumont. 



During the past winter I had occasion to study sever- 

 al of our cruciferous weeds aud in that connection a 

 study of the seeds and testa revealed some interesting 

 points, so that it has seemed wise to publish the results 

 of this work, though it is not complete with reference to 

 all the species described in Gray's Manual. The seeds 

 are so characteristic that our weedy species are easily 

 distinguished. 



Tlie seeds are round, flattened, oval, rough or smooth. 

 These characters can be made out easily in sections. Cot- 

 yledons flat, incumbent, —the back of one cotyledon 

 lying against the caulicle, or accumbent with the edges 

 of the cotyledons towards the caulicle, or longi- 

 tudinally plicate and partially enveloping the caulicle 

 or conduplicate in cross section (Mustard) or spirally 

 coiled in some cases, A section magnified shows that 

 the testa consists of two well-detined layers and some- 

 times of a third, which is much compressed. The cell- 

 walls of the outer portion are mucilaginous ; those of 

 the second thick-walled. The aleurone layer which 

 various authors have considered as belonging to the testa 

 is endosperm. Strasburger speaks* of the seed as 

 l»eing exalbuminous, and most systematic botanists so 

 speak of it in this way. In the sense that this term 

 was used by early systematic botanists this is cor- 

 rect. Humphrey has called attention to the aleurone 

 layer of seeds and the use of the term. Endosperm of 

 this character is found in the seed of many Legumino- 

 sa3 where it cannot be made out with the naked eye. 

 We have not thus far found it wanting in the order 

 LeguminosejB. This layer corresponds to that found in 



*Haii(lhoii;^ of Practical Bot., Eufjlish translatiou, HilUiouse, p. 339, 



