60 
year. The American Academy of Arts and Sciences at Boston, 
and the National Research Council at Washington, have granted 
us material assistance during the past two years. But we can not 
expect these grants-in-aid to be continued from year to year. The 
breeding of chestnut trees is a long-time project which may extend 
over a period of ten years or more in the future. If we could be 
assured of $1000 a year for that period, the chances of ultimate 
success would be much increased. Such an investment should 
yield most satisfactory returns to the donor. 
Acknowledgments —We take pleasure in acknowledging the con- 
tinued cordial cooperation of the Division of Forest Pathology of 
the U.S. Dept. of Agriculture with us in this problem; and we 
sincerely appreciate the many helpful letters, the specimens of 
nuts, and material assistance in other ways, from sources too nu- 
merous to mention here. These evidences are sufficient proof of 
the great public interest in this problem. 
Respectfully submitted, 
Artitur TH. GRaAvEs, 
Curator of Public Instruction. 
SYSTEMATIC BoTANy 
The Classification of Dicotyledons 
By ALFRED GUNDERSEN 
Studies continued during 1936 indicate clearly that there is one 
change in the Engler System for which the evidences appear to be 
especially convincing. Briefly, the groups of families represented 
by Cistus and Papaver should not be separated. That is, Engler’s 
groups, Parietales and Rhoedales, in America often called Violales 
and Papaverales, belong together. This is not saying that these 
groups are necessarily wholly natural ones; only that the system 
— 
of dicotyledons becomes greatly simplified by having these groups, 
characterized in the main by parietal placentation, adjacent. 
With my last year’s report was included an outline of seven 
systems of Dicotyledons (Sympetalae excepted) from 1824 to 
1925. It shows that in the De Candolle, Bentham and Hooker, 
Kichler, Warming, Wettstein, and Rendle systems, Papaver and 
