82 
AMERICAN JOURNAL OF BOTANY 
[Vol. 9 
diseases caused by Pestalozzia are concerned, that authors have, in general, 
used the names P. Guepini and P. palmarum interchangeably, perhaps 
preferring the former name for tea diseases and the latter for palm diseases, 
but no one has justified the latter usage either by cross-inoculation studies 
or by finding dependable morphological distinctions. 
We were unable to attempt cross-inoculation studies in the time at 
our disposal. Likewise, as advocates of biometrical methods in systematic 
work, we feel called upon to apologize for the scantiness of our statistical 
data, which were obtained as a by-product, as it were, of other work. We 
were in general unable to make more than twenty-five measurements of 
spore length, and an equal number of appendage length, in each generation. 
However, we have data for each strain covering from four to eight genera- 
tions. The material proved unusually favorable for statistical study because 
of the presence of two independent and easily measured variables, namely, 
spore length and appendage length. In the measurement of spore append- 
ages we chose at random one of the three appendages of each spore, by 
the simple expedient of measuring whichever appendage lay most perfectly 
in the focal plane. 
In spite of the fact that our procedure might have been improved, we 
feel that our general conclusion, that the nominal species in the Fungi 
Imperfecti can be resolved into a large number of quantitatively distinct 
strains, is well established. Greater refinement of method would merely 
have increased the significance of the differences between the more closely 
similar groups of strains, and could only have confirmed the conclusion that 
might have been drawn from purely theoretical considerations, namely, 
that in the Fungi Imperfecti, a group of non-sexual organisms, every 
hereditary modification, of whatever magnitude, must give rise to a pure 
line. These pure lines are too numerous and too incapable of precise 
recognition to be named. Consequently the concept of species must be 
arbitrary and ruled by practical considerations, the species including a 
large group of distinct strains. Any one strain is capable of precise defini- 
tion within certain limits, established by the degree of refinement of en- 
vironmental control and biometrical accuracy, but within these limits it 
is indistinguishable from other strains which theoretically may exist, but 
cannot be recognized. 
Our cultures were all grown upon Hevea-leaf agar with native brown 
sugar, the latter being the boiled-down sap of the sugar palm, Arenga 
saccharifera. Different lots of medium were fairly uniform. On the whole, 
the successive generations were subjected to no more environmental fluctua- 
tion than would probably obtain in the average laboratory, since tempera- 
ture conditions in Asahan are very uniform. No effort was made to keep 
the cultures at constant temperature. The strains showed physiological 
differences among themselves, indicated by the different lengths of time 
required for the production of spores and by the varying abundance of 
