OcTOBER 8, 1897. ] 
tubers were prevented from growing and 
numbers of flowers were formed. This re- 
sult he also looked upon as indicating a 
compensation of growth between the flowers 
and tubers. While we recognize Knight’s 
experiments as of great importance, yet he 
erred in his interpretation of the results of 
this supposed correlation ‘between the tu- 
bers and flowers, as V6chting has shown. 
By repeating Knight’s experiment, and also 
by growing shoots so that tubers would be 
prevented from developing, while at the 
same time the roots would be protected, 
flowers were obtained in the first case while 
they were not in the second, so that the 
compensation of growth, or correlation of 
growth, here exists between the vegetative 
portion of the plant and the flowers instead 
of between the production of tubers and 
flowers as Knight supposed. 
The theory of metamorphosis as expressed 
by Goethe and A. Braun (’51) and applied 
to the leaf regarded the leaf as a concept or 
idea. As Goebel (’80) points out, Braun 
did not look upon any one form as the 
typical one which through transformation 
had developed the various leaf forms, but 
each one represented a wave in the march 
of the successive billows of a metamorphosis, 
the shoot manifesting successive repetitions 
or renewals of growth each season, present- 
ing in order the ‘ niederblatter, laubblatter, 
hochblatter, kelchblatter, blumenblatter, 
staubblatter, fruchtblatter. Though it had 
been since suggested from time to time, as 
Goebel (780) remarks, that the foliage leaf 
must be regarded as the original one from 
which all the other forms had arisen 
(though at that time Goebel did not think 
this the correct view), no research, he says, 
had been carried on, not even in a single 
ease to determine this point. Goebel 
plainly showed in the case of Prunus padus 
that axillary buds which under normal 
conditions were formed one year with sey- 
eral bud scales could be made by artificial 
SCIENCE. 
541 
treatment to develop during the first year. 
This he accomplished by removing all the 
leaves from small trees in April, and in 
some cases also cutting away the terminal 
shoot. In these cases the axillary shoot, in- 
stead of developing a bud which remained 
dormant for one year as in normal cases, at 
once began to grow and developed well- 
formed shoots. Instead of the usual num- 
ber of bud scales, there were first two stip- 
ule-like outgrowths, and then fully ex- 
panded leaves were formed, so that in this 
case, he says, the metamorphosis of the leaf 
to bud scales was prevented. For this re- 
lation of bud scales to foliage leaves Goebel 
proposed the term ‘correlation of growth’ 
(780). In the case of Vicia faba removal 
of the lamina of the leaf of seedlings when 
it was very young caused the stipules to 
attain a large size and to perform the func- 
tion of the assimilating leaf. He here 
points out that experimentation aids us in 
interpreting certain morphological phe- 
nomena which otherwise might remain ob- 
secure. He cites the occasional occurrence 
(Moquin-Tandon) in the open of enlarged 
stipules of this plant which his experiment 
aids in interpreting. In the case of Lathy- 
rus aphaca the stipules are large and leaf- 
like, while ghe part which corresponds to 
the lamina of the leaf is in the form of a 
tendril, the correlation processes here hay- 
ing brought about the enlargement of the 
stipules as the lamina of the leaf became 
adapted to another function. Kronfeld 
(86, ’87) repeated some of Goebel’s ex- 
periments, obtaining the same results, and 
extended them to other plants (Pirus 
malus and Pisum sativum), while negative 
results attended some other experiments. 
Hildebrand, in some experiments on seed- 
lings and cuttings, found that external in- 
fluences affected the leaves, and in some 
cases, where the cotyledons were cut, foliage 
leaves appeared in place of the usual bud 
scales, and in Ozalis rubella removal of the 
