34 Dr. H. T. Brown and Mr. F. Escombe. [Jan. 9, 
twenty times the normal amount. We have made an exhaustive series of 
test-experiments on this point and have found by introducing a second 
Reiset apparatus “in series” that there is no falling-off of complete absorption 
in the first apparatus until from one-fifth to one-fourth of the alkali has 
been carbonated, and that even when this point is reached the falling-off of 
absorption is due to a virtual shortening of the column of alkaline solution by 
the excessive carbonation of its lowest section; for when the contents of the 
three “stages” of the tube are thoroughly mixed the column once more 
becomes completely absorbent. In actual practice only a very small fraction 
of the total amount of caustic alkali is carbonated, and providing the column 
of liquid is not less than 50 cm. in length the speed of the air-current is 
practically only limited by the danger attending the ejection of liquid from 
the tube. 
At the completion of an experiment, if the whole amount of alkali has to 
be submitted to the method of “double titration,” which is always the case 
with the smaller apparatus, the vertical tube is removed from the glass vessel, 
rinsed out with a little water free from CO,, and over the aperture is slipped 
an india-rubber cap furnished with a tubulure through which the end of the 
“ acid-burette” can be afterwards passed for titration. In using the larger 
apparatus it was found more convenient to titrate only an aliquot part of the 
solution; and here we are met with a practical difficulty which must be 
referred to somewhat in detail. During the passage through the Reiset’s 
2 
“tower” of large volumes of air only partially saturated with moisture, a 
certain amount of evaporation necessarily takes place, so that the final volume 
of the absorbing liquid is somewhat less than the initial volume. In addition 
to this it 1s necessary to use a little water to wash the upper part of the 
“tower,” owing to the breaking bubbles of the liquid spurting a little of the 
solution of soda on to the sides of the tube. The final volume of liquid can 
be determined from the difference in weight of the apparatus at the beginning 
and end of the experiment, and the known volume of the original alkaline 
solution taken. If we represent the original volume of the alkaline solution 
in cubic centimetres by V, and the difference between the initial and final 
weight in grammes of the charged apparatus by D, then V+D represents 
the final volume of the solution in cubic centimetres if no contraction of 
volume has taken place on adding the wash-water. Theoretically such a 
contraction undoubtedly takes place, but it cannot appreciably affect the 
result since it is the special merit of the method of titration employed that an 
error in the determination of the volumes at any stage of the experiment can 
only affect the final carbon dioxide estimation in the proportion which this 
