DISTRIBUTION OF NITROGEN 



95 



since temperature variations of the atmosphere affect the small 

 volumes of gas which have to be measured. 



(2) The deaminising bulb : 11-12 c.c. 



(3) The burette : 2 c.c. 



The quantities of reagents required are 10 c.c. of sodium nitrite 

 solution and 2-5 c.c. of glacial acetic acid for which the correction for 

 impurities amounts to 'O6-*I2 c.c. It is not necessary to have a 

 smaller Hempel pipette ; with the micro apparatus the permanganate 

 lasts for a considerable time. 



The solution to be analysed may be introduced into the burette 

 with a pipette and the burette is washed with six or seven drops of 

 water. 



The error in the estimation need not exceed '005 mgm. when 

 2 c.c. or less gas is measured ; with more gas 'Oi mgm. One- 

 fifth of the amount of solution required for the macro apparatus is used 

 in the micro apparatus. Not only is there an advantage economically 

 with reagents, but also the apparatus is less fragile. 



0*5 mgm. of amino acid can be analysed with an accuracy of 

 I per cent. The analysis is slightly more 

 rapid : at 1 5-20 four minutes' shaking suffice, 

 at 20-25 three minutes, above 25 two to 

 two and a half minutes. 



It is essential that the burettes be ac- 

 curate and the stopcocks be tight. The 

 burettes are tested by weighing the water 

 they deliver ; the stopcocks by submitting 

 them to a pressure of a column of water one 

 metre high. 



The apparatus is cleaned with a mixture 

 of potassium bichromate and sulphuric acid. 



Micro Apparatus ii. 

 The size of apparatus can be reduced 

 to one-half of the above, if the form of the 

 gas burette be modified [Van Slyke, 1915, 

 2]. This is shown in fig. 4. The zero- 

 point, instead of being placed at the bottom 

 of the stopcock, is located on a capillary 



FIG. 4. 



After J. Biol. Chem., 1915, 23, 

 408. 



which extends for a few millimetres below the tap. It permits 

 marking off the upper boundary of the gas volume with an error of 



