1915] 



COULTER — ORIGIN OF MONOCOTYLEDONY 181 



tissue and by the very obvious vascular relations. A cross- 

 section of this very specialized embryo is instructive (figs. 8 

 and 9). The large functional cotyledon is seen originating on 

 one side, embracing the vascular axis of the embryo and more 

 or less overlapping the other side, where in most grasses the 

 second cotyledon (epiblast) appears. Moreover, in the section 

 of the centrally placed plumule, with its succession of leaves, a 

 section of the stem tip may be seen, clearly representing the 

 axis of the embryo, with no suggestion of a lateral origin. A 

 transverse section through the cotyledonary plate (fig. 

 shows some tissue developed at the site of the missing cotyle- 

 don (not overlapped by the functioning cotyledon). This is 

 emphasized by the appearance of a mass of procambium at the 

 base of the protuberance, which in other grasses develops into 

 the epiblast. This procambium is distinctly a rudiment of a 

 former vascular connection. 



Some idea of the frequency with which the second cotyledon 

 appears among the grasses may be obtained from the excel- 

 lent work of Bruns on the grass embryo, published in 1882, 

 and from the work of Van Tieghem, already cited, published 

 in 1897. Bruns examined 82 genera, representing 12 tribes. 

 In 29 of these genera epiblasts were present, and the genera 

 represented 9 of the 12 tribes. The tribes in which no epiblasts 

 were found were Oryzeae, Agrostideae, and Aveneae. The 

 situation in the Agrostideae is noteworthy, for 13 genera were 

 examined, and no trace of an epiblast found. Festuceae may 

 be mentioned, for 20 of its genera were examined, and only 4 

 of them were found to possess epiblasts. Taking Bruns' re- 

 sults as a whole, they indicate that approximately 40 per cent 

 of the grasses still develop a second cotyledon to a stage that 

 enables it to be recognized under ordinary inspection as a 

 definite structure. 



The work of Van Tieghem included a somewhat wider range 

 of forms, 91 genera being examined, and 61 of these showed 

 epiblasts. This suggests that perhaps in as many as two- 

 thirds of the grasses a second cotyledon is more or less ob- 

 vious. In any event, it is certain that the grasses as a whole 

 exhibit a remarkable number of transition stages from dicoty- 



