144 
sized city filter plants. The cities in the 
neighborhood of Troy supply good illustra- 
tions of the different modern methods of water 
purification. 
Tue Bulletin of the American Mathematical 
Society states that during the Easter vacation 
of 1912 an extensive course in mathematics 
and physics for advanced teachers will be held 
at Gottingen under the direction of Professor 
F. Klein. 
Dr. CO. Rankin has been appointed pro- 
fessor of botany, and director of the Botanical 
Gardens at Copenhagen. 
Dr. Ginpert T. Morcan, assistant professor 
of chemistry at the Imperial College of Sci- 
ence and Technology, South Kensington, and 
junior hon. secretary of the Chemical Society, 
has been appointed to the chair of chemistry 
at the Royal College of Science, Dublin, va- 
cant by the retirement of Sir Walter Noel 
Hartley, F.R.S. 
Dr. TuHropor Boveri, professor of zoology 
at Wiirzburg, has been called to Freiburg. 
DISCUSSION AND CORRESPONDENCE 
NUMBER OF STUDENTS PER TEACHER 
To THE EprTor oF ScrENncE: It appears to me 
that the only correct way to determine the 
average number of students handled per 
teacher in any school is to divide the number 
of student hours per week by the number of 
teacher hours per week. 
For example, let there be 15 teachers and 
300 students. This does not mean that on 
the average one teacher instructs 20 students 
in a recitation or class. Suppose each student 
takes 15 hours per week, and that each teacher 
instructs only 12 hours per week. There are 
therefore 15 X 12 classes per week for 300 < 15 
students, since each student appears in 15 
classes. The average number of students in 
each class is therefore 
300 X 15 +15 K 1225. 
In general, therefore, the average number 
of students which each instructor has to handle 
in one recitation is the number of student 
hours divided by the number of teacher hours, 
in one week. 
The average number of hours per week re- 
SCIENCE 
[N.S. Vou. XXXV. No. 891 
quired of each student and each teacher, viz., 
the number of student-hours and teacher- 
hours per week divided respectively by the 
number of students and teachers, are also im- 
portant numbers in respect to the average 
work required of students and teachers. 
ArtHur §. Harnaway 
ROSE POLYTECHNIC INSTITUTE 
FUNDULUS AND FRESH WATER 
“ Fundulus and Fresh Water” in your issue 
of December 29, 1911, recalls some experi- 
ences I have had with these fishes in transfer- 
ring them from salt and brackish to fresh 
water. Fundulus heteroclitus may be so 
transferred more safely, the less degree of 
salinity there is in the water whence they 
were derived. Most of my specimens came 
from the Hackensack River and its creeks, 
varying from the saline Newark Bay to the 
almost entirely fresh water at Little Ferry 
and at the heads of the creeks. While very 
few of those transferred from salt water di- 
rectly to fresh survived the sudden change, an 
increasing number survived of those gradu- 
ally transferred in the course of a week or 
two, through a number of changes of water. 
My reeords show that such fishes lived from 
four to six months, up to two years; one lived 
over three years. I never succeeded, however, 
in making a successful transfer of the highly 
colored breeding males. 
Fundulus diaphanus, though known almost 
entirely as a fresh water species, when taken 
from salt water also offers difficulties in trans- 
ferring, thus showing that successful trans- 
fer in all cases appears to be a matter of very 
gradual accomplishment. 
Fundulus majalis I never succeeded in 
transferring, no doubt because of its being a 
purely marine species. 
Cyprinodon variegatus also can be gradu- 
ally accustomed to a change of water, but be- 
ing practically only an anadromous fish dur- 
ing breeding time does not very long survive. 
It is possible that these cyprinodonts being 
great rovers can ill bear small quarters and 
this may be one reason for their shortlivedness 
in captivity, as compared with the quieter 
eyprinids for instance. 
