TERMINOLOGY IN MONOCOTYLEDONS. 515 



in their terminology is admitted; and even in this respect 

 there is often more difference between the first and second empty 

 glume than between the upper empty glume and the first 

 flowering one ; and in some cases, as, for instance, in the common 

 Couch-grass (Agropyron repens) (PI. IX. fig. 5), the empty and 

 flowering glumes are precisely similar or very gradually diminish 

 in size from the outer empty to the uppermost flowering ones. A 

 particular empty glume in one spikelet may also correspond to a 

 flowering one in another spikelet of the same plant. Thus in the 

 Rye- Grass (Loliumperenne) (PI. IX. fig. 6), the spikelets are alter- 

 nately placed in one plane, right and left of the main axis of inflo- 

 rescence, the single empty glume of each spikelet being the lowest 

 and outer one, and therefore alternately the right-hand and the 

 left-hand one, whilst the second glume, next the axis of inflores- 

 cence, is the lowest flowering one. In the uppermost spikelet, 

 however, which, from the non-production of the main axis, is appa- 

 rently terminal, there are two empty glumes : and this is not owing 

 to the development of an additional outer glume ; for the lower of 

 the two empty ones is on the side it ought to be in the regular 

 alternation with the lower spikelets; but the second glume, which 

 in the lower spikelets encloses a flower, is in this subterminal one 

 empty. So in several Paniceae, the second or third glume, accord- 

 ing to the genus or species, has been observed sometimes to enclose 

 a rudimentary, or a male, or even a perfect flower, and at other 

 times to be quite empty without any consequent change in its 

 appearance. It has then been termed either one of the paleas of 

 the enclosed flower, or, when empty, a neutral flower. Now in 

 all other orders, whether Monocotyledonous or Dicotyledonous, 

 a neutral flower signifies one in which the floral axis with the 

 perianth (always regarded as part of the flower) are developed, 

 but the organs of both sexes are either sterile, rudimentary, or 

 deficient ; but in these so-called neutral flowers in Gramineae the 

 whole flower, perianth, bracteoles, axis and all, is deficient, and 



there remains nothing but the subtending bract,which is in no other 

 case regarded as a part of the flower, still less as a whole flower. 

 Thus in Panicum (PI. IX. fig. 3), according to the Kunthian ter- 

 minology, the first minute scale is a glume, the second, many times 

 larger, is alsoaglume, the third, often precisely similar to the second, 

 is not a glume but a flower, and the fourth, whether similar or more 

 or less dissimilar, is a part of a flower. In some Gramme® th 



are additional empty glumes, usually small and often different in 

 form, either immediately below the flowering ones, as in Antho- 



