1916 



Langdon; on Carbon Monoxide 



241 



viously discussed. The blood from the chick killed by the kelp gas was 

 not reduced by ammonium sulfide. 



The above qualitative tests were quite conclusive and are preliminary 

 to the discussion of quantitative results, which follow. 



In the early part of the work extra precautions were taken to pre- 

 vent loss of carbon dioxide. The method of collecting the gas for exam- 

 ination is illustrated by Fig. 2, the gas being forced out of the cavity and 

 into the analytical apparatus by mercury. Thus the gas came in contract 

 only with mercury while being transferred to the absorption apparatus. 



Since the "complete" analysis of the gas, previously discussed, did not 

 show hydrogen or hydrocarbons, the analytical device used for most of 

 the work was a Williams modified Orsat appartus, specially designed for 

 the determination of carbon dioxide, oxygen, and carbon monoxide. Re- 

 sults which checked to within two to three parts per thousand were ob- 

 tained. Throughout the work fresh carefully prepared reagents were 

 used, and the accuracy was tested daily by the analysis of known gases and 

 air. In all cases a 100 cc. sample of gas was analyzed. 



Table 2 gives the results of 12 analyses which are typical of those 



Table 2. Results of typical gas analyses, shoxving small carbon dioxide 

 content; gas collected over mercury. 



SAMPLES 



%CO 3.8 



%0' 18.8 



%CO» 0.0 



Table 3. Comparison of gas analyses at various times of day and night. 



•The data were taken from samples collected .July 4, 1916. The day was 

 bright except for an hour in the early afternoon. The tide was high ( + 8.7 ft.) 

 between 5 and 6 A. M., then low ( + 2.3 ft.) at noon and high ( + 11.2 ft.) again 

 between 8 and 9 P. M. Temperature of the water lOo c. 



