692 - REPORT OF COMMISSIONER OF FISH AND FISHERIES. 
With a shorter tube the divisions may be made proportionately shorter. We intend 
to have at least forty parts by weight of fine soda lime for every part of substance in 
the “mixture.” About 0.4 and 0.6 gramme of flesh (water free), or corresponding 
amounts of other materials, we find convenient. The importance of fine pulverization 
of substance, intimate mixture with soda lime, filling the tube compactly so as to 
have no channel, avoiding the shaking by which particles of the substance might 
be brought to the top of the soda lime, and heating until no charred material is left, 
is insisted upon. An anterior layer of coarse soda lime, 12 cm. long, has proven 
very satisfactory. The anterior layer should be well heated before the heat is 
applied to the mixture of soda lime and substance, and kept hot until the combustion 
is done. 
3. Heat and time of combustion . — A “low red” heat ordinarily suffices. Heating to 
bright redness brings danger of loss of ammonia by dissociation, though in our expe- 
rience, when the tube is closely packed and the operation not too slow, we have found 
practically no difficulty in getting all the nitrogen as ammonia even at high heat. But 
with a channel in the tube the loss by high heating has been considerable. 
Ordinarily, three-quarters of an hour is ample for the combustion, and an hour the 
extreme limit, according to the experience in this laboratory. Rapid combustion is 
less and long heating more dangerous than is frequently taught. With an ordinary 
Knop and Arendt (four bulb) bulb apparatus containing 10 cc. of acid solution, of 
which little over half was required to neutralize the ammonia, the ammonia was com- 
pletely absorbed even when the combustion lasted only 12 minutes. Serious loss 
may result from too long heating, especially if there be open space in the tube or the 
temperature is high. 
4. Determination of the ammonia . — In the combustion of ordinary protein compounds 
with the precautions stated above, practically all the nitrogen is converted into 
ammonia, and its determination by titration is easy. We find it well, however, to 
use concentrated solutions and to avoid excess of water in rinsing out the bulbs. The 
quantity of concentrated acid solution required is small and the tension in the com- 
bustion tube during the heating consequently slight, which is a convenience ; while, 
with the small quantity of solution in titrating, the color reaction is sharp and the 
determination easy and accurate. Freshly prepared cochineal solution is the most 
satisfactory indicator we have found. Very narrow burettes, in which 10 cc. occupy 
from 30 to 40 cm., have decided advantages for convenience and accuracy when con- 
centrated solutions are used. 
Accuracy of measurement df solutions . — The danger of error in measur- 
ing the standard acid and alkali is greater than is sometimes supposed. 
The results of our observations may be briefly summarized :* 
(1) Leaving temperature out of account, uniformity and accuracy of 
delivery depend upon the completeness with which the liquid is removed 
from the inner walls of the tube. The amount of adhering liquid is 
greater the larger the extent of the interior surface to be drained ; the 
greater the amount of dirt, grease, etc., on this surface ; the more rapid 
the outflow and the shorter the time allowed for the after flow, i. e ., 
for the adhering liquid to drain down. When the tubes are not clean 
and the delivery is rapid the amount of adhering liquid may be so large 
and so variable as to materially affect the measurements. But with 
clean, narrow tubes and fine jets the uniformity, and hence the accu- 
racy, are all that could be desired. 
* See article on Burettes and Pipettes by W, O, Atwater and C. D. Woods, Journal of 
Anal. Chem., vol. i, p. 373, 
