448 
J. ARTHUR HARRIS AND JOHN V. LAWRENCE 
In considering these exceptions it is important to note that the 
differences between parasite and host observed in this series as a whole 
are not large. The average differences for 42 determinations, regard- 
ing signs, is 0.154° while ^oi* the 37 cases in which the osmotic pressure 
of the parasite is greater than that of the host the average difference 
is only 0.203°. 
Working as we did under many difficulties the greatest exactness 
in the constants cannot be expected. In at least one of these excep- 
tions the difference between the freezing point lowering of the two 
organisms is so small that it may be due to purely experimental errors. 
Thus one would hardly assert that such a difference as 0.018° for 
Dendrophthora on Baccharis furnishes any valid proof that the osmotic 
pressure of the sap of these parasites is really lower than that of the 
leaves of their hosts. 
With regards to these exceptions the following biological factors 
should be pointed out. 
In the case of Collection 203, the value for the parasite is among 
the three smallest out of the 14 made for Phthirusa parvifolia. It 
may, therefore, be actually too small. The host tree had apparently 
been injured by fire recently. Old and young leaves of the host were 
taken. The young leaves gave only A = 0.931. Thus the parasite 
(A = 1. 1 81) shows a concentration distinctly higher than do the young 
leaves of the host which are drawing their water in competition with 
the old ones. 
The difference between the freezing point of D. cupressoides and 
its host, B. scoparia, is, as has already been pointed out, so slight 
that it might easily be due to an error in the determination of the 
freezing points alone. The constant for Baccharis is the highest 
secured for that species, which because of the hardness of its tissues 
presents some difficulty in the extraction of sap. Another point will 
be mentioned below. 
Thus of the 42 comparisons there seem to be only three, that is 
those numbered 447, 508 and 509 in which the evidence of a superior 
osmotic pressure of the leaves of the host can be given any weight at 
all. These three cases are particularly interesting. 
Note first of all that two of these three exceptions are leafless and 
rather thick stemmed shrubs, Dendrophthora gracilis and D. opuntioides. 
It is quite possible, as explained above, that in these cases there has 
