12 
addition, as formaldehyde exists in the air, in smoke, and elsewhere 
as an incidental and often unsuspected product. 
For the determination of the formaldeh 3 "de in strong solutions, the 
method most highl}^ recommended is that of Blank and Finkenbeiner 
(19). Legler’s volumetric method (20) is also in frequent use. 
For the determination of the formaldehyde content of weak solu 
tions Romijn’s (18) iodometric method is the most convenient and is 
trustworthy if the solution is known to be pure; other aldehydes and 
acetone are especially not allowable. In dilute solutions containing 
these or unknown substances in addition to the formaldehyde it is 
safer to use the cj^anide method (IT). 
The exact determination of small amounts of formaldeh 3 "de in 
admixture with other organic matter is alwa 3 "s difficult and often 
impossible, owing to chemical changes taking place between the 
organic matter and the formaldehyde or between the reagents 
employed for the formaldeh 3 "de and the organic matter. Colorimet- 
ric methods (23, 21, 25) furnish approximate results, and in the case 
of milk, at least, (26) a nearl 3 ^ constant fraction of the formaldeh 3 "de 
can be distilled over if exact methods are employed. 
In quantitative determinations of formaldeh 3 ule much care is needed 
on account of its volatilit 3 " and instabilit 3 ". In man 3 " of the best 
methods, also, unanticij^ated reactions are apt to vitiate the results 
unless directions are followed exacth"; hence it is recommended b 3 " 
Fresenius and Gruenhut that no result should be accepted without the 
concordance of at least two distinct methods (27). 
WORK BY PREVIOUS INVESTIGATORS. 
There has been an immense amount of work done by investigators 
upon the antiseptic and germicidal value of formalin. The following 
is a brief resume of all the principal articles. 
Trillat (28) showed that the following percentages of formaldehyde 
are necessary to prevent putrefaction of bouillon kept at 30° C. : 
1:50,000 delayed putrefaction very sensibly. 
1: 25,000 bouillon unaltered at the end of four days. 
1:12,000 remained unaltered after several weeks. 
Bouillon inoculated with bacillus aiithracis was rendered sterile by 1: 25,CC0. 
The putrefaction of 10 c. c. of bouillon inoculated with 10 gtts. human caliva was 
sensibly delayed by 1:50,000. 
1: 30,000 remained unaltered at the end of fourteen days. 
1: 1,000 killed all the organisms in less than two hours. 
Gelatine contaminated Avith sewage water containing 1,800,000 organisms to the 
c. c. showed no growth after being treated Avith 1: 20,000. 
1:1,000 killed all of the organisms after some hours exposure. 
Its power of preventing fermentation in milk and wine is veiy 
marked in ver 3 ^ small proportions, 1:4,000 in wine putting an end to 
fermentation. 
