256 
AGNES CHASE 
spicata few herbarium specimens reveal these cleistogenes, but if the 
curly tufts so common in late autumn on sterile knolls and rocky hill- 
sides be examined they are invariably (in my experience) found with 
cleistogenes in the broken culm or in the internodes remaining in the 
tuft. I surmise they are a regular rather than an occasional method 
of reproduction. 
The chasmogamous spikelet of Cottea pappophoroides is a highly 
specialized one, the florets being deeply 5- to 7-cleft and awned, and 
partly hidden in copious white hai'rs. The cleistogene is usually 
solitary, without glumes, and consists of a single floret with a length- 
FiG. 3. Cottea pappophor- Fig. 4. Pappophorum Fig. 5. MuUenhergia 
aides. Ordinary spikelet WrightiL Ordinary spike- niicrosperma. Ordinary 
and cleistogene and grain of let and cleistogene and spikelet, cleistogene en- 
each- X 5. grain of each-X 5- "^^^'^^ ^^^tending leaf, 
cleistogene removed, and 
grain of each — X 5. 
ened rachilla joint bearing a minute rudiment of a second floret. 
The lemma is sparingly silky villous and minutely awned or awnless 
(fig. 3; note the relative size and shape of the grains). The prophyl- 
lum is delicate and split as it is in Danthonia. In a single case a spike 
of three one-flowered spikelets was found. In these the second glume 
was developed, and this infolded the floret and rachilla joint. 
In Pappophorum Wrightii the chasmogamous spikelet bears a 
4 
