ANATOMY AND PHYLOGENETIC POSITION OF BETULACEAE 423 
where they grow through space and thus fertihze the egg by way of 
the micropyle. 
Another feature worthy of mention is the way in which these 
plants are polUnated. Among the Dicotyledons there are two chief 
agencies for carrying this out. All of the higher members of the 
group are pollinated by insects. Such plants usually possess well 
developed floral envelopes. On the other hand the Betulaceae, 
Fagaceae, and Casuarinaceae together with other families of the 
lower Dicotyledons and all of the Gymnosperms are not pollinated in 
this way. PoUination in these takes place through the agency of the 
wind. This has long been recognized as a characteristic common to 
the lower famiUes of the Angiosperms. If these forms are reduced to 
any great extent, they would be expected to show some trace of insect 
pollination. An instance may be mentioned of a family which is 
even now placed very low because of its apparently simple floral 
structure but which is pollinated by insects. This family is that of 
the Salicaceae belonging to the cohort Salicales. Miss Holden (191 2) 
has worked with these and, since the principles which she has used are 
the same as those which are being used here, it will not be amiss to 
give in some detail a summary of her results. Through recent work 
it has been shown that the presence of wood parenchyma and its 
distribution are very definite things and hence of great value in deter- 
mining the evolutionary relationships of a plant. In the lowest Gymno- 
sperms no wood parenchyma is present. When it first occurs, it appears 
only at the end of the annual ring (terminal) and is evidently developed 
in correlation to the alternation between summer and winter in connec- 
tion with storage of food material. The next step, and one which 
is common to the higher Gymnosperms and the lower Angiosperms, 
is the diffusion of the wood parenchyma throughout the entire annual 
ring. Among the higher Angiosperms the wood parenchyma is more 
or less localized around the vessels. In Salix the parenchyma is 
terminal. This condition by itself is deceptive, since not even the 
higher Gymnosperms are so characterized. Again the rays throughout 
our common species of Salicaceae tend to be uniseriate like those of 
the Conifers. Because of this structure and by reason of the appar- 
ently simple flower this family has been placed at or near the bottom 
of the Dicotyledons. . 
Miss Holden in making her study used exactly the same principles 
which have led to such excellent results among the Gymnosperms. 
