THE FLOEAL ENVELOPES IN GRAMINE^'; AND CYPKRACE^. 71 



Fig. 17. — Kobreda caricina : the outer bract turned back and exposing' tlie 

 ochrea or spatlial bract. 



Gramine^. 

 I now go ou to record, in the Order 

 Gramine(v, a few instances of abnormal de- 



velopment which are relevant to our present \ 

 subject, and shall end my paper by giving \ 

 the conclusions which I think may fairly be \^ 

 drawn from the facts and reasoning brought 

 forward. 



Botanists are generally agreed that the 

 terms "pale," "fertile glume," "upper and 

 lower barren glumes," are more in accord- 

 ance with the homology of the spikelet of 

 grasses than the terms " inner or upper 

 pale," "outer or lower pale," and "upper 

 and outer" or "lower and inner glumes"; 1'^ 



and all deviation from the ordinary form or 



arrangement of parts, or, in other words, all instances of abnormal 

 development are of special interest if they support these views. 



I have, in my notes, the record of the occurrence, in 1873, in 

 Hampshire, of a specimen of Lolium perenne, or common rye-grass, 

 in which the upper, outer, single, and usually barren glume of one 

 of the spikelets was transformed into a fertile one by the develop- 

 ment of a bifid pale (exactly similar to the pale of a normal 

 fertile floret) alternate with, i.e. subtending, the glume. The 

 ovary was, to all appearance, perfect, but there was only one 

 stamen. 



Both Babington and Koch call the suppressed glume in Lolium 

 the " upper" glume, while Kunth calls it rightly the "inner" or 

 " lower" glume. It is evidently the lower barren glume which is 

 wanting, as may be jn-oved by its position in Festuca loUacca, and 

 also by the alternation of the florets, as well as by the homology 

 of the spikelet. Now this lower glume being the inner one, there 

 niust be, normally, a suppressed outer and subtending leaf or 

 bract at the base of each spikelet, and the evidence of its normal 

 position and suppression may be distinctly seen, by the presence of 

 a raised border extending transversely more than half way round 

 the axis of the spikelet (it is seen more faintly at the base of the 

 upper spikelets), and this bract sometimes becomes developed more 

 or less. 



The presence of a membranous border, in a similar position, in 

 many grasses, is very evident, especially in Cynusurus ciistatu.s, and 

 it occasionally becomes considerably developed, as in Scsleria 

 carulea. I have also specimens of Serrafalcus arvensis (fig. 18), and 

 S. commutiitus, in which the bract is developed as a leafy setaceous 

 appendage. 



I once gathered a specimen of Festuca loliacea in which the 

 upper barren glume of one of the upper spikelets was developed 

 into a perfect flower, as in the case of Loliuw. In some spikelets 

 of F. luliacea the inner or lower barren glume is wanting, or is 



