238 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1914. 
stallation, occupying an unusually large floor space and expensive 
buildings. 
The general tendency is now toward cells which can be used in very 
large units, which can be housed economically, and of which the gen- 
eral cost of maintenance and renewal is small; some of the modern 
types of diaphragm cells are now successfully operating with 3,000 
to 5,000 amperes per cell. 
As to the possible future improvements in electrolytic alkali cells, 
we should mention that in some types the current efficiencies have 
practically reached their maximum, and average ampere efficiencies 
as high as 95 to 97 per cent have been obtained in continuous practice. 
The main difficulty is to reinforce these favorable results by the use 
of lower voltage without making the units unnecessarily bulky or 
expensive in construction or in maintenance, all factors which soon 
outweigh any intended saving of electric current. 
Here, more than in any other branch of chemical engineering, it is 
easy enough to determine how ‘“‘good”’ a cell is on a limited trial, but 
it takes expensive, long-continuous use on a full commercial scale, 
running uninterruptedly day and night for years, to find out how 
‘“‘had”’ it is for real commercial practice. 
In relation to the electrolytic alkali industry a great mistake is 
frequently committed by considering the question of power as para- 
mount; true enough, cheap power is very important, almost essen- 
tial, but certainly it is not everything. There have been cases where 
it was found much cheaper in the end to pay almost double for electric 
current in a certain locality than in another site not far distant from 
the first, for the simple reason that the cheaper power supply was 
hampered by frequent interruptions and expensive disturbances, 
which more than offset any possible saving in cost of power. 
In further corroboration, it is well known that some of the most 
successful electrolytic soda manufacturers have found it to their ad- 
vantage to sacrifice power by running their cells at decidedly higher 
voltage than is strictly necessary—which simply means consuming 
more power—and this in order to be able to use higher current densi- 
ties, thereby increasing considerably the output of the same size units, 
and thus economizing on the general cost of plant operation. Here 
is one of the ever-recurring instances in chemical manufacturing where 
it becomes more advantageous to sacrifice apparent theoretical effi- 
ciency in favor of industrial expediency. 
All this does not diminish the fact that the larger electrochemical 
industries can only thrive where cheap power is available. 
Modern progress of electrical engineering has given us the means to 
utilize so-called natural powers; until now, however, we have only 
availed ourselves of the water power developed from rivers, lakes, 
and waterfalls. As far as large electric power generation is concerned, 
