476 



Mr. E. C. Grey. Decomposition of 



strain derived from it, but did not agglutinate the strain CF. And, likewise, 

 the serum obtained by inoculating a rabbit with the normal strain CF 

 agglutinated the normal strain CF, and the selected strain derived from it 

 (CF selected) up to a dilution of 1 : 25600, but did not produce the slightest 

 agglutination with the normal or selected strain No. CI.* This may be seen 

 from Table II. 



Analytical Technique. 



The methods of analysis described by Harden (1901) have been for the 

 most part closely followed ; certain slight modifications, however, have been 

 introduced, which may be described here. 



Volatile Acids. — In the steam distillate which is used for the determina- 

 tion of formic and acetic acid, the formic acid has been determined by the 

 formation of mercurous chloride, and the acetic acid obtained by subtracting 

 the amount of formic acid so found from the total acids determined previously 

 by titration of the whole distillate with alkali, using phenolphthalein as 

 indicator. Two errors are introduced here due to the presence of small 

 amounts of carbonic acid and lactic acid in the distillate. The carbonic acid 

 has, therefore, been estimated by barium hydroxide and the lactic acid by 

 Eyffel's method. This estimation of lactic acid in the distillate becomes of 

 importance when the amount of acetic acid is small. 



The distillation to obtain the volatile acids was carried out in two stages. 

 The first fraction was obtained without admitting steam, measured about 

 400 c.c, and contained the alcohol and part of the volatile acid. This 

 fraction was titrated with standard baryta solution. A slight excess of 

 baryta was then added, and the solution distilled with a fractionating column 

 in order to remove the alcohol. The residual fluid now contained a granular 

 precipitate of barium carbonate, which was removed by rapid filtration and 

 titrated at the boiling point with N/10 H 2 S0 4 . The barium hydroxide 

 corresponding to the barium carbonate was deducted from that required to 

 neutralise the first distillate. In this way an accurate correction may be 

 made for the carbon dioxide dissolved in the distillate. 



After removal of the first 400 c.c. steam was admitted, and the distillation 

 continued until 100 c.c. of the distillate required only 01-0'2 c.c. of normal 

 alkali for neutralisation. The total steam distillate usually measured about 

 2500 c.c. The distillate was neutralised with potash, united with the first 

 fraction, and the whole evaporated to dryness. The residue was dissolved 

 in 100 c.c. of water, and an aliquot portion used for the determination of 



* The agglutinating sera were kindly prepared for me by Dr. J. A. Arkwright of this 

 Institute, to whom my best thanks are due. 



