CHEMICAL COMPOSITION OF FOOD-FISHES. 
691 
Though it is desirable to avoid coloration and turbidity of the acid solution, these 
do not necessarily imply incomplete ammonification, nor does their absence prove 
perfect combustion. With proper care to insure contact between soda lime and 
substance, we have almost never found the solution so colored as to seriously interfere 
with titration. 
(9) Ammonia may be dissociated and nitrogen lost by either too high heating or by 
conducting the operation so slowly as to leave the ammonia exposed for a long time 
to heat. It seems probable that the presence of water vapor* as of other gases, would 
tend to prevent dissociation of the ammonia, and that the danger of long heating 
may be partly due to reduced supply of moisture from the anterior layer of soda 
lime after the latter has been heated for some time. 
(10) A vacant space in the tube (channel as ordinarily recommended) may cause 
serious loss. This loss is greater the higher the temperature and the longer the time 
of combustion. It is probably due not only to incomplete ammonification of distilla- 
tion products through lack of contact with the soda lime, but also to dissociation of 
ammonia. With the channel the flow of the gases is slower and they are exposed to 
heat longer than when the tube is packed full. Add the possible lack of water vapor 
when the heating is long continued, and the loss by dissociation is very clearly 
explained. When the tube is closely packed, the flow of gases reasonably fast, and 
the operation conducted at a temperature sufficient to heat the tube only to dull 
redness, there appears to be no considerable loss by dissociation, even with a long 
anterior layer (20 or 30 cm.) of soda lime. 
Concerning reagents, apparatus, and manipulation, a few words will suffice. 
1. Soda lime . — The soda lime made by mixing one part of ordinary caustic soda 
with two and a half parts of quicklime by the process described,* costs very little 
for materials and labor, and serves the purpose very satisfactorily. In sifting it is 
conveniently divided into a finer portion to be mixed with the substance, and into 
coarser particles to be used for the anterior layer. It bears heating without fusing 
so much as to leave any considerable open space in the tube if closely packed at the 
outset. Varying proportions of soda lime, from one part to two and one-half parts 
of lime to each part by weight of soda, have made no difference in the results of the 
analyses. We have obtained equally good results with the mixture of sodium car- 
bonate and slaked lime as described by Johnson, and see no reason why slaked lime 
as recommended by him should not be generally efficacious as it has proven in the 
cases cited by him and in those tried by ourselves. Our reason for adhering to the 
ordinary soda lime has been the impression that, by filling the anterior portion of 
the tube with coarse particles of the rather difficultly fusible material, more complete 
contact is insured between nitrogenous distillation products and the heated water 
vapor from the soda lime. The old theory that enough soda should be mixed with 
the lime to make the mixture easily fusible does not stand the test of experience. 
In testing the purity of soda lime by sugar, -as is sometimes recommended, there is 
danger of error both from the presence of nitrogen in the sugar and from formation 
of acid distillation products. 
2. Tubes and charging . — For ordinary combustions, tubes of from 35 to . 40 cm. in 
length do very well. The method of charging the tube upon which we have gradually 
settled, after numerous trials with tubes of different lengths and charged in different 
ways, is explained in the accompanying tabular statement : 
Centimeters. 
Length of tube 40 
Asbestos and fine soda lime 4 
Mixture, fine soda lime and substance 16 
Rinsings, fine soda lime 4 
Anterior layer, coarse soda lime 12 
Asbestos, open space, plug 4 
* Amer. Chem. Journal, 9, 312. 
