Shaw.— The Seedling S true hire of Araucaria Bidivillii. 331 
In Araucaria Bidivillii the same variation in internal structure is 
associated with a similar abnormality in the external morphology. In 
other words, numerous protoxylem groups prevail at the base of a tuberous 
hypocotyl. Moreover, the number of these groups shows considerable 
variability, seedlings with heptarch, hexarch, or pentarch structure being 
equally common. It was not possible to trace any relationship between 
the degree of tuberosity of the hypocotyl and the variation in the number 
of protoxylem groups in the root structure at its base, although this 
variability may be looked upon as evidence that the structure at this point 
is the result of the operation of biologic factors. In the Cupressineae 
an increase in the number of bundles in the hypocotyl is caused by an 
increase in number of cotyledons. Here, however, certain bundles behave 
as though coming from half-cotyledons, and there is no deviation from the 
normal character of the root structure. But the occurrence of heptarch, 
hexarch, or pentarch structure in A. Bidwillii obviously cannot be corre- 
lated with any splitting of cotyledons. 
Some seedlings, in which the main tap-root was rather shorter than the 
average for the specimens, did not attain to diarch structure, remaining 
triarch, or even pentarch in the case of specimens beginning with heptarch 
structure at the base of the hypocotyl. It takes, therefore, some time for 
the seedling to obliterate the influence of its swollen hypocotyl and to 
reduce its structure to the normal. In very young seedlings, of course, the 
root would show no such reduction. This circumstance perhaps accounts 
for the statement by Borzi (1) that the root was always pentarch. Another 
tuberous species, A. imbricata , shows, as far as can be gathered from the 
existing description, an essential similarity to A. Bidivillii , although 
numerous protoxylems in the base of the hypocotyl do not seem to be 
so distinct as in the latter species. 
Thus the relationship between the seedling structure of A. Bidivillii 
and the usual type shown by seedling Conifers seems to be exactly paral- 
leled by that subsisting between Eranthis and the remaining Ranunculaceae. 
It is difficult to avoid the conclusion that the similar peculiarity in habit is 
directly connected with the corresponding abnormalities in structure. 
While in any one natural group of plants the transition phenomena may 
show a unity in the general method, yet the actual number of separate 
xylem groups in the root seems to depend upon the particular character of 
the individual. Any variation from the normal number of protoxylem 
groups in an aberrant species seems not likely to be of serious taxonomic 
value, but is more probably to be correlated with the habit of the seedling. 
The study of seedling anatomy from a biological standpoint probably offers 
a more fruitful subject for investigation than is likely to be afforded by the 
attempts made to found far-reaching conclusions upon the vagaries of 
seedling structure. 
