I9I4] Sergius Morgulis 439 



ADDENDUM 



Since the foregoing was written (March) I have had occasion to 

 install, in the Woods Hole Laboratory of the Bureau of Fisheries, 

 a small apparatus for the determination of carbon dioxid. The 

 work with this apparatus is still in progress and the results will be 

 published later, but I wish to add here a brief account of some of 

 my experience in handling this apparatus, which supports the above 

 criticism of the technic of Benedict and Cathcart's investigation. 



When I first put up the apparatus, it seemed quite sufficient, for 

 the complete removal of moisture, to bubble the air through two 

 connected wash-bottles containing sulfuric acid, one of which was 

 half-full of glass beads (3 mm. diam.). This seemed specially 

 effective because the Ventilation was relatively slow (probably less 

 than one hundredth as rapid as that in the Benedict-Cathcart appa- 

 ratus) and, although the air was saturated with water vapor, the 

 surface of contact with the absorbent was very extensive. 



In the early blank experiments with this apparatus, an increase in 

 weight of the soda-lime and calcium chlorid tubes was observed that 

 was very hard to explain. Various possibilities were tested and at 

 last, as it seemed certain that the increase in weight must have been 

 due to retention of water, a third sulfuric acid wash-bottle was 

 added to the series. The effect was unmistakable, and the increase 

 in weight was at once shifted from the first to the third decimal 

 place. This experience was rather striking and convinced me of the 

 need of very strict control at that point of the apparatus. I there- 

 fore inserted between the last sulfuric acid wash-bottle and the 

 soda-lime tubes, a U-tube with glass beads submerged in a little 

 sulfuric acid in the bend. This tube is weighed, now, before and 

 after each experiment; and constancy in its weight is the only 

 guaranty that the accuracy of the technic is not affected from that 

 source. 



In spite of the increased labor, the results with this new arrange- 



ment and experimental procedure are most gratifying. In later 



successful blank experiments the change in weight was found to be 



very insignificant. The improved apparatus can be employed with 



complete reliance upon the accuracy of the results obtained with it. 



Laboratory of the Bureau of Fisheries, 

 Woods Hole, Mass. 



