410 
at any time by moving the furniture from 
one room to another. Another administra- 
tive advantage will be the ease with which 
the laboratory may be enlarged by adding 
a few more unit rooms, each of which will 
be as perfectly suited to all the varied re- 
quirements of the laboratory as the rooms 
already in use. Such enlargements may 
be indefinitely repeated as long as the build- 
ing space holds out. 
The advantages of instruction are those 
which ensue from having the students di- 
vided into small sections. These were 
considered in the previous article, reference 
to which has been made above. I venture 
to repeat what was there said. There are 
two principal alternatives between which 
we may choose—the adoption of large labo- 
ratories or of a series of small laboratories. 
The choice is between keeping the students 
in large classes and dividing them into 
small sections. Personally I can advocate 
only the latter choice and must plead for it 
very strongly. In a large laboratory with 
75 or 100 or more students, the noise and 
confusion are necessarily great, and the de- 
tailed supervision of the work is extremely 
difficult. If the students are subdivided 
into small sections, these and other diffi- 
culties at once vanish, and if each section 
can be assigned a separate room it may be 
put under the charge of a special instructor, 
who shall be personally responsible for the 
work of that section. Thus each assistant 
can be given a certain independence. The 
sense of opportunity with the accompany- 
ing responsibility will tend to improve the 
quality of histeaching. He can be directed 
to carry his section over a certain part of 
the subject in a given time and left free to 
accomplish the result. If the section is 
small enough the work can be interrupted 
for an explanation, a direction or a quiz, 
and the students will ask more and better 
questions than in the large laboratory, 
where the many listeners embarrass them, 
SCIENCE. 
[N. S. Vou. XIII. No. 324. 
and where they may not find always the 
same instructor at hand. Finally, the in- 
structor can learn the personal qualities 
and needs of the men in a small section 
and establish personal relations with the 
individuals. It should never be forgotten 
that such personal relations are most im- 
portant factors in efficient teaching. It 
might be urged against the position here 
taken that the class could be divided into 
sections in a large room, but my experience 
convinces me that such a scheme is utterly 
impracticable. The advantages of moderate 
sized rooms for all advanced work and for 
investigations of every kind are so com- 
monly admitted, that their justification by 
argument is uncalled for. The unit rooms 
should be therefore of suitable size for these 
purposes also. 
Assuming now that the unit system is 
desirable, we pass on to the consideration 
of what may be the best dimensions. The 
problem resolves itself into three questions : 
First, What is the best unit space to allow 
for each student ? 
Second, What isthe most convenient max- 
imum number of students to assign to a 
single room? 
Third, What additional space must be pro- 
vided in each room for passages, sinks, 
shelves, etc., so as to permit it to be fully 
equipped for the class work ? 
Let us now consider these questions in 
the order given. 
First. The unit space required for each 
student. In the previous article twenty 
square feet was suggested as a preliminary 
estimate, or an area four by five feet. Since 
then a number of laboratories have been 
measured, and the relative advantages of 
various unit areas considered. It is of 
course desirable on the one hand to reduce 
the space reserved for each student in order 
to reduce the total size of the building re- 
quired, and on the other hand to give each 
student ample room to carry on the prac- 
