J. H. Long — Polarization of Tartrate Solutions. 353 



ture and amount of the change is characteristic of certain 

 groups of salts. These points are shown below. 



At the outset I prepared a quantity of pure Rochelle salt, 

 sufficient for all experiments contemplated, and determined the 

 rotation with five solutions containing exactly 5, 15, 25, 35 and 

 45 gm in 100 cc , measured at 20° C. 



The results of these tests are* given in the table, the specific 

 rotation being calculated from the formula, 



10' a 



M = LTC7 

 in which a is the observed angle of rotation, L the length of 

 the tube containing the solution in millimeters, and C the con- 

 centration of the solution, that is, the number of grams in 100 cc . 



T 



L 



C 



Sp. Gr. (2£) 



a 



[a] 



20° 



200 mm 



5 



1-0261 



2°'214 



22°-14 



20° 



200 mm 



15 



1-0739 



6°'650 



22°-16 



20° 



200 mra 



25 



1-1202 



1L°*'058 



22°12 



20° 



200 mm 



35 



1-1655 



15°-493 



22°-13 



20° 



200 mm 



45 



1-2100 



19°'854 



22°-06 



The specific rotation appears to be practically constant with 

 the varying concentration, and the change in the angle of rota- 

 tion by variation of temperature is inappreciable as I found by 

 observing several of the solutions when heated to 35°. Krecke 

 found (Arch. JVeerland., vii, 202), with a Eochelle salt solution 

 containing 20 gm in 100 cc , the specific rotation [a] D = 22° 42 at 

 25° C. For a solution containing 10*771 gm of the anhydrous 

 salt in 100 cc Landolt found (Berichte, vi, 1*073), [a] D = 29°*67, 

 which is equivalent to 22° '09 for the crystals. 



This is practically the same as I obtained in the table above, 

 and as I found it in numerous other tests carefully conducted I 

 am inclined to think it is the true specific rotation. In what 

 follows I will take 22°*1 as the constant at 20°. 



For each one of the following tests I took 20 gm of the 

 Rochelle salt and dissolved it in a narrow-necked flask grad- 

 uated to hold 100 cc at 20°. Then the amount of the inactive 

 salt was added and water enough to nearly reach the mark. 

 After shaking, the flask was brought to the right temperature 

 by suspending it in a large vessel of water kept at 20°. A little 

 distilled water having the same temperature was added and this 

 was continued, until, after shaking, the mark was just reached. 

 The limit of error here is about one-twentieth of one per cent. 



The solutions so made were then used in the polariscope. In 

 all the experiments here given the temperature was kept at 20° 

 and the 200 mm tube was employed. 



The following table embraces the results obtained : 



Am. Jour. Sci.— Third Series, Vol. XXXVI, No. 215.— Nov., 1888. 

 23 



