16 
stream of air could thus be sent through the retort im order to take up the 
carbon dioxide and lead it over into the absorption tubes. Å wash-bottle was 
inserted before the retort and filled with soda-lime to cleanse the stream of 
air of carbon dioxide. From the retort the stream of air passed through a U 
tube in an asbestos-covered oxen; this tube contained granulated copper dioxide 
(after Messinger, Kjeldahl uses here dioxide of mercury). The object of 
this copper dioxide, which is greatly heated up during the separation, is to oxidize 
certain products which are not completely oxidized to carbon dioxide in the retort. 
Behind the oven is inserted a wash-bottle with concentrated sulphurie acid to dry 
the stream of air, which then passes through two weighed U tubes with sodium 
carbonate which takes up the formed carbon dioxide in the stream of air. 
The separation was made with an addition of 10 gm. of bichromate of 
potassium and 30 cm.? concentrated sulphuric acid to the substance. The whole 
was heated up slightly until gas was given off. The duration of the awalysis was 
1//, hours. 
Before the quantity of carbon is determined, all the carbon dioxide from 
carbonates must naturally be removed first of all. This is done by covering ths 
weighed substance with a strong solution of phosphorie acid and letting it dry 
at 60—709, 
From the found quantity of carbon dioxide the quantity of carbon is 
calculated and this is again calculated in percentage of the bottom-sample im 
dry condition. 
To test with what degree of accuracy the determination of carbon could be 
made by this method, I weighed 251 mgm. of pure cane-sugar, which according 
to calculation should yield 387 mgm. carbon dioxide; 365 mgm. carbonic acid was 
obtained, thus ca. 5"/, too little. When we remember that it is the relative and 
not so much the absolute values we want in these investigations, an error of this 
dimension cannot be said to have any importance. 
The determination of the nitrogen was made according to Kjeldahl's 
method. 
The result of the analyses was as follows: 
1. North Sea, 3 miles west of Agger; depth 16 meters. Two bottom- 
samples were taken, the one consisting of fairly pure sand, the other with somewhat 
more clay. Equal parts of the samples were mixed together. This sample con- 
tained 0.34/, C and 0.027%, N. The quotient NS == 20: 
2. Nissum Bredning, off Røiensø Odde; depth 6 meters. The bottom- 
sample contained a good deal of sand. It consisted of 0.58%/, C and 0.045 %%, N. 
The quotient = — 124. 
3. Kaas Bredning. The bottom in Kaas Bredning greatly resembles 
the bottom in Thisted Bredning, which was described in detail above. The 
bottom-sample used for analysis was taken in the middle of the Bredning in 6 
meters depth. It contained 1.58, C and 0.142 %, N. The quotient s == IL 
4. Kaas Bredning. Another bottom-sample was taken in Kaas Bred- 
ning off Kaas Hoved at 11 meters depth. As was to be expected, this sample 
contained a little more carbon than sample 3, namely 1.98 9%,. 
