ANAEROBIC RESPIRATION 73 



posed between the container and the apparatus for the absorp- 

 tion of carbon dioxide. Wash bottles with the exit tube wound 

 in a coil are very practical. Direct determinations show that 

 by the use of these flasks even a minimum loss of alcohol is 

 prevented.^ At the close of each experiment, the experimental 

 material and the contents of the wash bottle are placed in a 

 large distilling flask, the container and the wash bottle are 

 rinsed with water, considerably more water is added, and the 

 whole is distilled until about half of the liquid has passed over 

 into the receiver. If the distillate is acid it is neutralized bv 



ml 



the addition of CaCOs. If, on the contrary, the first distillate 

 is alkaline (which happens only in exceptional cases), it is 

 rectified by the addition of phosphoric or tartaric acid. If a 

 cloudy distillate is obtained the flask is placed at a slope for 

 the next distillation, as for steam distillations. The use of dis- 

 tillation heads such as are used in Kjeldahl determinations of 

 nitrogen, also give good service. The volume of the alcohol 

 solution is reduced to about 50 cc. by further distillations. 

 This solution must be perfectly clear and neutral in reaction. 

 If a qualitative demonstration of alcohol is desired, the 

 solution is concentrated as far as possible by repeated distilla- 

 tions. The following are recommended as qualitative tests: 



1. The appearance of oily streaks in the upper part of the 

 condenser at the beginning of the distillation. After some 

 practice the presence of ethyl alcohol may be recognized with 

 certainty by the appearance of these oily streaks.^ This test 

 is not very sensitive. Traces of alcohol are not detected by it. 



2. The iodoform reaction of A. Muntz.^ 10 cc. of the clear 

 distillate are mixed with 2 g. sodium carbonate and o.i g. sub- 

 Hmcd iodine and the mixture is warmed on the water bath at 

 60° until the iodine is dissolved. After cooling there separate 

 out the characteristic crystals of iodoform, which may be 

 easily recognized under the microscope. The iodoform test 

 is really very sensitive but not entirely reliable. It also gives 

 a positive reaction with various other substances, among which 



1 Kostytschew, S. und P. Eliasberg. Z. f. physiol. Chem. iii : idi. 1920. 



2 Pasteur, L. Compt. rend. 52: 1260. 1861; cf. also Hansen, A. Meddel. fra Carlsberg 



ab. I : 175. 1881. 



3 Muntz, A. Ann. de chim. et de physique (V) 13: 543- 1878. 



