98 MR. R. Н, COMPTON : AN INVESTIGATION OF THE 
intermediate-high transition with a diameter of 1:9 mm. Here it seems 
necessary to call to our aid a hereditary factor, high and intermediate-high 
transitions being the general rule in the Lote. 
What, then, is the nature of the apparent correlation between the length 
of the hypocotyl and the level of transition? It appears that this correlation 
is secondary, and follows from the fact that in the majority of seedlings there 
exists a rough proportionality between length and diameter. There appears 
to be an average type of form for the epigeal Leguminous seedling: if the 
diameter be increased the length tends to be increased in the same ratio. 
We can demonstrate this tendency by a comparison of columns III. and IV. 
in the table, when it is seen that in each class of seedlings the length of the 
hypocotyl is 14-20 times the diameter at its base, the average ratio for the 
120 epigeal seedlings being L/2R=17. This being so, it is clear that a 
correlation of transition-level with diameter will also appear as a correlation 
with length. 
It follows that the figures quoted in column VI. of the table merely 
emphasise those in columns 1V. and V. without adding anything new to the 
discussion. 
In a few cases, however, the length of the hypocotyl may have a direct 
effect upon the classification of transition-levels itself ; this is especially the 
case in middle transitions. If the hypocotyl be short there is less room for 
the transition than if long, and in consequence the changes may apparently 
take place at a higher level, relatively to the collet and cotyledonary node. 
(See, for instance, Scorpiurus sulcata and 5. vermiculatus, p. 85, and certain 
Genistee, p. 80.) This cause has probably contributed to the relative 
vagueness of the classification of middle transitions. 
THe LEVEL oF TRANSITION AND THE MATURE HABIT. 
The Summary List reveals the fact that all those species which show the 
tree-habit have low or intermediate-low transitions: out of 20 epigeal 
species of trees (“shrubs or small trees” excluded) 14 show a low, 6 an 
intermediate-low transition. On the other hand, the herbaceous habit appears 
to be associated with a high transition, though not so strictly: out of 55 
epigeal herbs 28 showed a high transition, 6 an intermediate-high, 9 inter- 
mediate, 3 intermediate-low, and 9 low (of which 7 are Phaseolez). 
There is thus some foundation for the belief that the mature habit is cor- 
related with the level of transition. But in view of the connection which 
exists between the transition-level and the diameter of the hypocotyl (p. 97), 
it appears probable that the correlation with habit is not an example of 
direct cause and effect. Asa matter of fact, it will be shown later that this 
