AXILLARY CLEISTOGENES IN SOME AMERICAN GRASSES 257 
single perfect floret and two or three sterile florets, the lemmas of each 
being split into nine spreading plumose awns, all the awns together 
forming a feathery, pappus-like crown. The cleistogene, which is 
sometimes so rotund as to split the sheath in which it is borne, consists 
of a floret without glumes, and a rachilla joint with a minute rudiment 
of a second floret. The lemma is more or less split into a few closely 
appressed lobes (fig. 4; note the relative size of the grains). The 
prophyllum is thin in texture and is usually not split. In two cases 
the cleistogene was infolded in the sheath of a reduced leaf, so judged 
from its position opposite the prophyllum and from the clearly evident 
differentiation into sheath and blade. In one case a spike with three 
one-flowered spikelets was found. In two species of the eastern hemi- 
sphere cleistogenes were found, namely Pappophorum boreale from 
Transbaical, Siberia, and Pappophorum brachystachyum from the 
Algerian Sahara. 
In Muhlenbergia microsperma the chasmogamous spikelet is very 
small, has minute glumes and a single long-awned floret. The cleisto- 
gene is usually awned, without glumes, and much more turgid than is 
the chasmogene (fig. 5; note the relative size of the grains). It is 
tightly folded in the swollen, spongy-indurate base of the sheath of a 
reduced leaf, which sometimes has a minute blade and ligule. In a 
few cases a tiny raceme was found with three or four spikelets with 
glumes like those of the chasmogenes. The cleistogenes are produced 
in abundance at the lower nodes, the subtending sheaths being pushed 
open by their bulk. (In old plants the numerous little cornucopias 
may be easily seen at the base.) The subtending prophyllum is 
membranaceous and remains attached to the nodes. Unlike those of 
Danthonia and the others, the nodes of M. microsperma do not disjoint, 
but the cleistogenes themselves readily fall at maturity. In Dan- 
thonia and the rest, the cleistogenes are permanently enclosed in the 
sheath together with the internode of the culm. (I surmise that the 
grains germinate within the sheath and push root and shoot through 
the internerves, but this has not been proved by experiment.) 
In all the species in which these cleistogenes have been studied 
they are found to be more variable than are the chasmogenes of the 
same species. Only a limited study can be made in the herbarium, 
and as yet only Triplasis, Danthonia, and Muhlenbergia microsperma 
have been studied in the field. Notes from observers would be grate- 
fully received. Since with relatively little study so many species have 
