164 BOTANICAL GAZETTE [Marca | 
great development in the types under consideration, and by 
researches bearing upon their relation to the nutrition of the 
structures in the embryo-sac.*5 
Whether the striking resemblances in structure between the — 
antipodal cells of Sparganium and those of certain grasses indi 
cate any direct relationship between the Sparganiacez and Grami: 
nee, remains to be seen; but in connection with the undoubted 
similarities in the structure of the embryo, it is by no meals. 
improbable that these two supposedly isolated orders of the 
monocotyledons may really prove to be related. 
That two such widely separated orders as Aracee and Com: 
posite should show analogies in those structures which may vey 
properly be assumed to be of importance as indicating relation 
ships, seems very surprising. It must be remembered, howevth 
that the Composite seem to have retained a very primitive Pt 
of ovule, 7. ¢., the solitary terminal ovule such as appears 10 
be the most primitive type known in the angiosperms. Ths 
same type also occurs in many aroids, and is found again in the 
simpler Centrosperme among the dicotyledons, é. g., Polygom 
cee. In the Piperinee, which show a number of curious 7 
gies with the Aracee, the same solitary axial ovule occurs. # 
is not impossible that a further study of Piperinee may § 
that in them we have a point of contact between monocot” 
dons and dicotyledons. If it should be shown that the sig 
ovule of the Composite is a primitive rather than @ secontay | 
condition, the occurrence of what seems to be a primitive 'P* | 
of embryo-sac in these most specialized of all angiosperms 
be comprehensible. tho 
Of course with our present very meager knowledge bot” 
the simpler types of dicotyledons and of the aroids, “a 
tive assertions as to the possible relationships of thesé gn 
are out of the question. That these groups deserve moe | 
ful study than has yet been given them, admits of little qu 7 
LELAND STANFORD UNIVERSITY. 
5 See especially Westermaier’s paper already referred to. 
