Nov., 1907.] 
Life History of Cornus Florida. 
1 9l 
front slopes, and below them new flood plains. Most of the 
terraces unless defended by rock ledges are subsequently cut 
out, the flood plain is replaced by a new lower one, and gradually 
but surely the old valley floor is more or less completelv uncov¬ 
ered and the old valley is restored similar to the form which was 
filled by the accumulating sediments. 
The whole cycle of aggradation and degradation constitutes 
but a little epicycle in the greater round of topographic evolution 
to which the hard rock lands must submit—a little pleasure trip 
which the stream takes while it rests from its great work of 
baseleveling a region. So short-lived are the forms due to these 
processes that the “eternal hills’’ do not change much during the 
entire period of the terrace stage and many epicycles may come 
and go and leave no record while one turn of the wheel of physi¬ 
ographic development is being made. 
CONTRIBUTION TO THE LIFE HISTORY OF CORNUS 
FLORIDA* 
William Clifford Morse. 
This study of the Flowering Dogwood was undertaken at the 
suggestion of Professor John H. Schaffner and the first material 
was collected September 26, 1905. When this material was 
ready for study, it was found that the flower-buds for the next 
year had already reached an advanced stage of development. 
Microspores were already formed and the ovule was far developed. 
Nevertheless, material was taken at intervals of a week during 
the fall, monthly during the winter, and weekly again during the 
spring. Through the kindness of Mr. Robert A. Young material 
was collected, from June until September, 1906, during the 
writer’s absence from Columbus. To him the author takes this 
opportunity of expressing his thanks. 
Schaffner’s weaker chrom-acetic acid solution was employed 
as a killing fluid. Before placing the head of flower-buds in it, 
however, the four bracts, which during the winter are tightly 
folded over them, and in the spring form the conspicuous invol- 
ucure, were removed in order that the solution could better 
penetrate the buds. Dehydrating and imbedding were per¬ 
formed in the usual manner. Sections were cut from 8 to IS 
microns in thickness. The vast majority, however, were 10 or 
12 microns. Analin safranin was first used as a stain and after¬ 
wards Delafield’s Haematoxvlin. The latter proved to be much 
the better and the best results were obtained by overstaining and 
then clearing for a long period in acid alcohol. 
* Contribution from the Botanical laboratory of the Ohio State University No. XXXI. 
