BROMUS INTERRUPTUS DRUCE. 19 



to me, in that single short stiff pedicels arise, alternately right and 

 left of the main rachis, each bearing at its extremity 3-5 sessile (or 

 in some cases shortly stalked) spikelets= To this fact is due the 

 peculiar and strikingly interrupted and compact appearance of the 

 whole inflorescence, which is made up of two rows of clustered 

 groups of 3-5 spikelets. This peculiar feature does not obtain in 

 its nearest allies, since in them 4-5 slender pedicels of various 

 lengths arise at the same level on the rachis, each bearing one or 

 two, rarely more, spikelets ; hence the inflorescence in these 

 species is more loosely continuous. As will be seen, the alliance of 

 B. interruptus is essentially with B. mollis, since the larger or inner 

 glume extends half way to the apex of the sixth floret (the third on 

 the same side), whereas in B. racemosus and B. commutatus it 

 reaches only to the fourth flower (the second on the same side). 

 The texture and pubescence too of the spikelets of B. intemiptus 

 are similar to those of B. mollia. The character which at once 

 separates B. interruptus from B. mollis, B. commutatus, B. racemosus, 

 &c., is to be found in the upper pale (the inner palea of Parnell), 

 wliich is uniformli/ split to the base, and is much shorter than the 

 outer or lower pale." 



The spht palea cannot, as may be suggested, proceed from an 

 accidental rupture during the growth of the fruit, as examination 

 shows that the pales are split from the early flowering stage. The 

 following note, added by Mr. C. B. Clarke to Mr. Druce's paper as 

 printed by the Linnean Society, shows that this character is of 

 exceptional interest : — 



"In all the examples of B. interruptus Druce, the upper pale is, 

 even in the young flower, divided to the base or at least yg^'^^s of 

 its length into two subequal lanceolar lobes ; each of these has the 

 green nerve down its middle (not exactly median), and the two 

 lobes stand soon rather divaricately apart. There is nothing in 

 any other species of Bromus approaching this ; and no such 

 complete division of the upper pale is well established in the whole 

 order. Dr. Stapf has lately examined into the few alleged cases of 

 the occurrence of a split upper pale in grasses ; and he cannot find 

 in verifying these any case i)arallel to that of Bromus interruptus ; 

 the split is either only partial, hardly half-way down, or it is 

 mechanical, i.e. does not exist in the young flower. . . . 



"Dr. Stapf argued, when Mr. Druce's paper was read, that 

 Bromus interruptus Druce must be treated as a monstrosity and 

 could in no case be made a new species. It may iudeed be main- 

 tained that the character of the completely bifid upper pale is 

 either generic or monstrous, one or the other, and cannot bft 

 specific. The remarkable uniformity with which the upper pale is 

 split to the base in every flower, in every specimen yet got from 

 diverse localities, may be held to negative for the present the view 

 that B. interruptus is a monstrosity. Of all the innumerable 

 species proposed as split-offs from Bromus vudlis, there is no one so 

 well worth a specific name as B. interruptus Druce, and no one to 

 be compared with it in morphologic interest." 



Bromus interruptus was first found by Mr. Druce in Berkshire in 



c 2 



