346 BOTANICAL GAZETTE [NOVEMBER 
results expected. The data for each hole were noted down sepa- 
rately, and the total was figured up when the preparation was 
finished. For the most part, for the reasons already stated, 16-holed 
plates were used, and usually 10 holes in each plate were examined. 
The basis of the interpretation of the figures thus obtained lay 
in the assumption that the average number of hyphae in each of the 
4 classes, considering a large number of holes, and supposing that 
no attractive or repellent forces existed, should, on the doctrine of 
averages, be approximately the same, that is, 25 per cent of the 
whole number; or, to put it in general terms, as many hyphae 
should point in one direction as another. That this assumption 
is legitimate is shown by the test count described later. Of the 
whole number, then, 25 per cent would constitute the normal 
number for each class, and any considerable deviation from this 
would indicate a reaction to a disturbing force. 
Since the B and D regions are of equal volume and equally 
subjected to repellent or attractive forces, they may be left out 
of consideration except in so far as the number occurring in them 
goes to make up the total for each hole, and thus influences the 
final percentage of reaction. In order to determine this percentage, 
then, we are concerned first of all with the A and C classes. Let 
us assume that an attractive force exists, derived from a substance 
on the other side of the mica plate, which is causing the hyphae 
to a certain extent to grow toward the holes. Since the number of 
the A class will then be greater than the normal number, we 
must conclude that those in excess of one-fourth have been attracted 
into this class. If the attractive force is not a very strong one, as 
_ indicated by a number not greatly above the normal, we may be 
fairly certain that the increment has been derived from those which 
would normally point right or left, that is, those of the B and D 
class. The number of A’s in excess of the normal, however, does 
not give us a complete measure of the attraction, for in all probabil- 
ity some hyphae originally in the C class have turned into the B or D 
direction, and the remaining C’s will therefore be fewer than one- 
fourth. The difference between the remaining C’s and the normal 
- number would then represent those which have turned from the C 
direction into B or D. By adding this number to those in excess 
