METHOD OF STUDY 137 
identify them, as far as possible, with previously described species, 
paying especial attention to the historical interpretation of the 
latter. 
One great lack in the study of such forms as those under con- 
sideration, is the want of correlation of observations. Here, more 
than in higher forms, different phases of the same plant are likely 
to be identified as different species ; again, because of a supposed 
resemblance between different forms, or, because two or more 
really diverse filaments are found growing together, they are asso- 
ciated under one name, without any proof of connection. 
In order to obviate this difficulty, and to place species defini- 
tion upon a firmer basis, we have made as thorough a study as 
possible, considering the large number of forms treated, of life 
histories. Our rule has been to make collections repeatedly from 
the same station, or rather from several stations, as well as to 
follow such collections by laboratory observation. 
Extended culture is a difficult matter in the case of algae like 
these which generally grow in rapidly running water. Such 
plants usually develop differently ina stagnant culture. To fit up 
a laboratory with a sufficient number of tanks supplied with fresh 
water would be expensive and not always feasible. A method, 
which has proved to be of value in certain cases, might profitably 
be employed further. We have sometimes placed fresh specimens 
in glass cylinders (e. g., large bottles with the bottoms knocked 
out), covered both ends with thin muslin, and anchored them 
near the surface in a running brook. Frequent examinations of 
such cultures may easily be made. Some plants do not take 
kindly to even this sort of captivity, perhaps because the change 
in the composition of the water is not beneficial ; others may be 
successfully kept until after important observations have been 
made. Many species have refused to disclose their reproductive 
processes even though carefully watched, but some advance in 
this direction has been made. 
As a rule, new species have not been proposed except in cases 
where repeated collections from the same stations have admitted 
of no doubt regarding the validity of the form in question. In one 
or two instances very marked forms have been described as new 
upon the strength of only one collection. Many specimens ap- 
