1910] CURRENT LITERATURE 313 
material, with special reference to the developmental history of the parts. By 
means of dissections and microtome sections of spikelets in various stages of 
growth, he arrives at what are represented to be definite conclusions concerning 
certain species in twenty-five genera. Among his results the following are of 
special interest: 
rom a phylogenetic standpoint he considers the grasses to have descended 
from lily-like plants in which there were six perianth divisions in two circles, six 
stamens, and a tricarpellate pistil. Among existing grass genera he places Strep- 
tochaeta as most primitive. In this genus there are three stigmas, six stamens, 
three inner and two outer organs which the author interprets as representing the 
inner perianth and the posterior segments of the outer perianth, all in the axil 
of the flowering glume or lemma. In the majority of grasses the floret, borne in 
the axil of the lemma, consists of a two-keeled palea, three stamens, two lodicules, 
and a pistil with a two-parted style. Hacket and others have considered the 
palea to represent the prophyllum, an organ found at the base of a branch, between 
that and the main stem. ScHUSTER states, however, that the palea is formed by 
the union of the two posterior segments of the outer perianth circle of the primitive 
flower. The inner perianth circle is well developed in Streptochaeta, but is usually 
reduced to small organs, the lodicules, which commonly function as spreaders 
during anthesis. The swelling of the lodicules pries open the lemma, thus expos- 
ing the stamens and stigmas. Usually there are present only two lodicules, the 
Posterior being absent, but in bamboos and some other grasses all three are present. 
In the more primitive forms, such as bamboos, there are six stamens, but in most 
8tasses three are suppressed, while more rarely only one or two are present. 
accounts for I t f certai genera, such as Pariana and Luzuola, 
by the splitting up of the original six stamens. The pistil consists of three carpels, 
but the anterior carpel usually fails to produce a style branch. 
Recently the morphology of the glumes of Elymus and Hordeum were investi- 
gated by ScHENcK,'s who considered the glumes to be developed from lateral 
branches at the base of the spikelet. ScHusTER controverts this, and states that 
the lower or outer glume originates as a single organ, but soon divides into two 
portions which stand side by side below the spikelet, the upper glume being sup- 
Pressed. In certain species of Elymus, such as E. arenarius, the glumes are in 
nearly normal position, for which reason the author follows HocHsTetTER*® in 
Separating this group under the genus name Leymus. The author further states 
that in the end spikelets of the spikes in Elymus and Hordeum, the glumes develop 
normally, that is, both are present and ehtire. 
Concerning viviparous grasses, it was found that the bulblet-bearing character 
Was more or less constantly inherited, but that by culture in a dry and nitrogen- 
poor substratum, these forms tended to revert to the normal seed-bearing state.— 
- S. Hitcucock. 
*S ENGLER’s Jahrb. 40:97-113. 1907. 
© Flora 31:118. 1848, 
