172 The Ottawa Naturalist. [Dec. 



would not recognize the epiblast as an independent leaf, was 

 because he found no vascular system in it; however, in Avena 

 sativa Didrichsen observed the epiblast to possess a very regular 

 lobation corresponding with the presence of several mestome- 

 strands. So far as coficerns the internode. which Van Tieghem 

 has declared to show the structure of simply a node, I must say, 

 that in Coix for instance, the structure is verv different from 

 that of a node, but identical with that of an internode, and 

 especially of a subterranean. In order to settle this question, 

 whether this stem-like portion, by Celakovsky called "mesocotyl," 

 be a node or an internode it is necessary to examine the internal 

 structure in a larger number of genera, when it is fully matured; 

 in Coix it is an internode, but it ma}^ not be constantlv so in 

 GraminecB and Cyperacece. Furthermore the presence of a bud 

 in the axil of the coleoptile, or according to Warming, in the 

 axil of the second leaf succeeding the scutellum, speaks in favor 

 of defining this leaf as independent of epiblast and scutellum. 

 Such axillary buds have been observed in a number of genera of 

 Hordecs; beside that Van Tieghem observed them in Avena, 

 and Bruns in Bambusa. The explanation offered by Warming 

 seems so simple and readily to be understood, that it is undoubt- 

 edly the most natural. 



In re\-ising these various views we notice that there is one 

 point, however, in w^hich all these authors agree, namely, that 

 the Graniincce have only one cotyledon. We shall see now that 

 there were formerly some authors who attributed two cotyledons 

 to this family, the scutellum and epiblast, thus the Graminecs 

 should possess one large and one small cotvledon; these authors 

 were Mirbel, Poiteau and Turpin. Recentlv Van Tieghem (1897) 

 has abandoned his former theory, and adheres now to the views 

 of these authors; in his new system he thus removes the 

 GraminecB from the other Monocotyledones , and places them 

 among his "'Inseminees." How untenable this classification is 

 has been shown by Celakovsky, who calls attention to a fact, 

 known long since, that in some genera oi Graminece the seed is not 

 grown together with the pericarp, but is free as in Eleusine, 

 Sporobohis, Crypsis and Heleochloa, and these genera should 

 consequently in accordance with Van Tieghem 's system be 

 separated from the other Graminecs, and referred to his 

 "Semin^es." 



The last type of monocotyledonous seedlings which may be 

 described here is exhibited by Peltandra undulata, Raf. (Fig. 18). 

 In this plant the fruit is a berry with a thin, almost black pericarp, 

 and contains mostly a single seed with noendosperm. The seed 

 is globular, surrounded by a tenacious jelly which, according to 

 Baillon and Engler, is the transformed exterior integument of 



