SEEDLING STRUCTURE IN THE LEGUMINOS.E. 97 
calculations are made *. The average length of hypocotyl for each class is 
given in column III. In the next two columns are given the average 
diameters at base and summit of hypocotyl. The numbers refer to milli- 
metres in columns IL.-V. In column VI. the results of a further calculation 
are exhibited, the object of which is to obtain a rough estimate of the volume 
in cubie millimetres of the average hypocotyl in each class ; the assumption 
being made that the hypocotyl is cylindrical. 
It is seen at once that the order of lowness of transition is also the order 
of magnitude of length, diameter, and volume of the hypocotyl. The lower 
the transition the greater the average dimensions of the hypocotyl. The 
parallelism is exact in all respects (save for a trifling discrepancy in 
column V.), but the results for volume given in column VI. are especially 
striking. 
In view of these results there can be no doubt that the level of transition 
is intimately connected with the absolute size of the hypocotyl. "While such 
a conclusion may be only partially justified in the case of the three “middle” 
transitions, it is amply proven in the three main grades of levei. 
Further, it is evident which of the two related characters must be regarded 
as cause and which as effect. The vascular system admittedly responds to 
change of external size and form, but it is almost inconceivable that structure 
should act as a cause determining size. 
We must therefore ask whether it is the length, the diameter, or the bulk 
of the hypocotyl which regulates the level of the anatomical transition. 
There is much to show that it is mainly the diameter (or more precisely 
the area of eross-section) which determines the level of transition. An 
inspection of the data in the Summary List reveals the fact that the length 
of hypocotyl for a given level of transition fluctuates on the whole within 
wider limits than the diameter. Then, again, it is clear that the length of the 
hypocotyl can have no direct bearing on the cases of very low transition com- 
pleted below the collet. As a matter of fact, we find low transitions in 
seedlings whose hy pocotyls vary in length from 5 to 110 mm. Further, the 
phenomena of transition are, so often associated with the region of the 
collet where a dilation of the axis occurs, that a direct connection between 
them and the diameter may be confidently assumed. The impression given 
in studying such cases is that the stele takes the opportunity afforded by 
the increased diameter to acquire a pith and expand. Where the diameter 
is small the stele perforce remains root-like and without pith until it actually 
divides in half toi supply the cotyledons. There are certain exceptions to 
this correlation, however, which are difficult to explain. ‘The most striking 
of these is, perhaps, the case of Lotus Tetragonolobus, where we find an 
* When two numbers appear, it implies that the first number refers to the species whose 
length, the second to those whose diameter, was estimated. 
LINN. JOURN.—BOTANY, VOL. XLI. H 
