6 
F. F. Blackman 
striking to find no family worthy to come in Class I or Class II, 
while most of them come at the tail in Classes IV and V. Here we 
seem to have a definite shifting of a biochemical character, possibly 
early in the separation of the two morphological types. Isolated 
exceptions occur to this general relation, as for example, Hydro- 
charts among the Monocotyledons which condenses its sugar to 
starch in great abundance. Both types of leaf, the starch-leaf and 
the sugar-leaf, seem to be quite adequate, physiologically, and 
neither suggests any real biological advantage over the other. 
This sugar-leaf and starch-leaf distinction can, however, be 
analysed a stage deeper, so that we believe we know what it turns 
upon. It is not due to any lack, in Monocotyledons, of the agent 
which determines the condensation of sugar to starch, but to the 
high critical concentration of sugar that is needed to start starch- 
formation. If sugar can be got into the cells in sufficient concentra¬ 
tion, then starch is visibly formed in leaves of the sugar-leaf type 
also. This can be done quite simply by floating cut pieces of leaf 
upon strong enough sugar solution in the dark. In 1898 Winkler 
showed that with ordinary starch-leaves, 0-2 to 0-5 per cent, sugar 
is the critical concentration. The moss Mnium gives the lowest value 
of 0*05 per cent. For sugar-leaves we have values like 15 per cent., 
while the leaves of the sugar-cane are found to require 18 per cent, 
sugar. Only one plant has such a high value that it has resisted all 
attempts to force it into starch-formation, and that is the Onion, 
which can be got to store nothing but sugar however it is treated. 
This * critical concentration ’ theory is supported by all workers 
on the subject. It indicates that there is not a simple equilibrium- 
relation between starch and sugar; and that, as Lundegardh has 
urged, starch-formation is a complex happening still requiring a good 
deal of quantitative investigation. One of the fascinating points 
of departure from simple chemical expectation is, that cane-sugar 
is a better sugar for artificial starch-formation than either maltose 
or glucose. 
We judge then that Monocotyledons are characterized by a high 
critical sugar-concentration generally, while Dicotyledons show a 
low one. This distinction seems to be merely one of degree and not 
of kind, so that it is hard to say to what sharply defined protoplasmic 
character it may be attributed. The distribution of this character 
over families is quite like that of succulence, showing here a wide 
sweep over a large number of related genera and there an isolated 
example. 
