S. L. Penfield — Methods for Determination of Water. 33 



in the tube by means of a fine wire, bent into a corkscrew coil 

 at the end. The results are satisfactory and are as follows : 



CuSo 4 . 5H 2 taken. 



H 2 calculated. 



H 2 found, 



Error. 



1. '2458 



•0887 



•0885 



•0002 



2. -3532 



•1274 



•1270 



•0004 



3. -3613 



•1304 



•1296 



•0008 



4. -3779 



•1363 



•1359 



•0004 



5. 1-7587 



•6347 



•6343 



•0004 



PeS0 4 . 7 H 2 taken. 



6. -1138 -0516 '0516 '0000 



7. -2150 '0975 -0978 -0003 + 



It frequently happens that water is to be determined in com- 

 pounds which give off carbon-dioxide on ignition, as for ex- 

 ample in rock analysis where calcite is present, and to test the 

 application of the method in such cases experiments have been 

 made with pure, recrystallized bicarbonate of potash, which 

 yields 8*99 per cent of H 2 and 22 per cent of C0 2 on igni- 

 tion, and also with gypsum mixed with calcite. It is evident 

 that carbon-dioxide will displace the air in the tube and must 

 be removed before weighing. This may conveniently be 

 accomplished by holding the tube at an inclination of about 

 40°, with the cap removed and open end down, so that the 

 heavier carbon-dioxide can flow out. At first the tube dimin- 

 ishes rapidly in weight, but after three hours the loss becomes 

 almost constant and amounts to very nearly 0'0003 grams for 

 every hour that the tube is left open. A correction therefore 

 of that amount must be made. Moreover, the carbon-dioxide 

 which passes out of the tube will carry water vapor with it, 

 but if the amount of the gas that is liberated is known a cor- 

 rection may be made. Assuming that the gas is saturated 

 with water vapor, and passes out of the tube at a mean baro- 

 metric pressure of 760 mm and a temperature of 20° C, 1 gram 

 of it would carry off "0096 grams of water. In the following- 

 series of experiments the tubes were left open in an inclined 

 position for three or more hours, and in addition to the time 

 correction, one was made for the water carried off by the 

 escaping C0 2 , which amounts to "0021 gr. for every gram of 

 HKCO, taken. 









Corrected 





HKC0 3 taken. 



H 2 calculated. 



H 2 weighed. 



weight. 



Error. 



1. -2694 



•0242 



•0225 



•0242 



•oooo 



2. -4641 



■0417 



•0395 



•0418 



•0001 + 



3. -5064 



■0455 



•0438 



•0458 



•0003 + 



4. 1-0055 



•0904 



•0869 



•0902 



•0002 — 



5. 1-0089 



•0907 



•0881 



•0912 



•0005+ 



6. 1-5109 



•1357 



•1312 



•1357 



•oooo 



7. 2-0000 



•1798 



•1747 



•1796 



•0002 — 



8. 2-7690 



•2489 



•2393 



•2463 



•0026- 



An. Jour. Sci— Third Series, Vol. XLVIII, No. 283.— July, 1894. 



