240 NINETEENTH REPORT. 



highest development of the group. In and of itself this arrangement is 

 perhaps almost as logical as any other. The difficulty is that it 

 emphasizes the differences between the grasses and other groups and 

 makes this family stand far removed from and without any good con- 

 nections with other groups. As a result most systematists have been led 

 to look upon the grasses either as standing very low in the Monocoty- 

 ledoneae or perhaps even as having arisen independently from some 

 ancestral Protoangiosperm. 



Hackel made a careful study of the structure of the grass spikelet and 

 flower and came to certain conclusions as to the homology of their parts. 

 According to him the flower is strictly achlamydeous, enclosed between 

 two bracts, the palea and lemma (using Hitchcock's terminology), one or 

 several such units being arranged either terminally or laterally (and then 

 alternately) on the rachilla, the whole being subtended by two glumes 

 and being known as the spikelet. The lodicules are looked upon by 

 Hackel as special organs to be considered as bractlets and not homo- 

 logous to any part of the perianth. The pistil in his opinion is mono- 

 carpellary, with a single basal ovule. The stamens are usually in one 

 whorl of three, rarely in two such whorls. 



This idea of the flower was not satisfactory and other botanists began 

 to look upon the lodicules as representing part of the perianth, most 

 likely two petals because of their position alternating with the outer 

 whorl of stamens. Miss E. R. Walker, working in Nebraska, demon- 

 strated by following its development from its first inception that tlie 

 grass pistil is really tricarpellary witli usually two stigmas and with the 

 ovule attached to the third carpel which bears no stigma. In rice, various 

 bamboos and a few other genera all three carpels have well developed 

 stigmas. Julius Schuster, working under the direction of Goebel on some 

 peculiar tropical grasses showed that the lodicules (which are usually 

 two in number, but three in some bamboos) represent true petals and 

 that the palea represents two sepals, mostly united (but then with the 

 keels of each forming the two keels characteristic of the palea) rarely 

 separate, and in some genera three in number. The usually missing 

 sepal is the one next to the subtending lemma, the missing petal being 

 next to the palea (two united sepals). 



Thus we see that the ordinary grass flower consists of a tricarpellary 

 pistil, one or two whorls of three stamens each, an incomplete whorl of 

 petals (lodicules) and an incomplete whorl of sepals (united to form the 

 palea), this flower being seated in the axil of a single bract (the lemma). 

 The type of flower just described will be recognized by botanists who 

 know the Monocotyledoneae as being of the general Lily type. In the 

 Lilies (using this term in the widest' sense to include the whole order 



