668 



SCIENCE 



[N.S. Vol. XXV. No. 643 



idea of the newest instruments is that of re- 

 versal so as to eliminate constant errors. 

 Diagrams were exhibited showing the method 

 of determination and the magnitude of errors 

 of graduation of certain English, French and 

 American observatory circles. The mechan- 

 ical process of graduating these circles was 

 briefly described. The relative merits of the 

 circles of different sizes used were discussed 

 and it was stated that the tendency is towards 

 smaller circles, one of the advantages of which 

 is possibly that the smaller circles are easier 

 to cast and to graduate than the large ones 

 with which some observatories had been equip- 

 ped. It was stated, as a general conclusion, 

 that the division errors of the circles men- 

 tioned by the speaker, and shown on the dia- 

 grams exhibited, are smaller than one would 

 expect them to be. E. L. Faris, 



Secretary 



THE AMERICAN CHEMICAL SOCIETY. NEW 

 YORK SECTION 



The sixth regular meeting of the session of 

 1906-'07 was held at the Chemist's Club, 108 

 West 55th Street, on April 5. 



Dr. C. A. Doremus presented a review of 

 the life of M. Berthelot and showed some in- 

 teresting pictures and memoirs. Dr. C. S. 

 Palmer added a few remarks regarding 

 Berthelot's work and finally the assembly rose 

 out of respect to his memory. 



The following papers were presented: 

 On the Danger of Over-specialization: L. H. 



Baekeland. 

 Some New Double Phosphates of Chromium: 



L. J. Cohen. 



1. The addition of a slight excess of diam- 

 monium phosphate to a hydrochloric acid 

 solution of ferric chloride precipitates a 

 double phosphate of iron of the formula 

 NH.H^O/FePO, which is perfectly white, 

 soluble in mineral acids and readily hydro- 

 lyzed by water and ammonia; prolonged boil- 

 ing with hot ammonia dissolves the salt with 

 the formation of a reddish brown solution 

 from which 95 per cent, alcohol precipitates 

 a basic double phosphate. On ignition, the 

 double salt decomposes, giving off ammonia 

 and water. 



2. When aluminum chloride is used instead 

 of iron, a corresponding double phosphate 

 forms of the formula NH,H^0,-A1P0, hav- 

 ing the same solubilities as the iron salt, 

 except towards alkalies; in the latter it dis- 

 solves completely, behaving like aluminum 

 phosphate. 



3. With slightly acid solutions of chromium 

 chloride diammonium phosphate precipitates 

 a double salt of the formula (NHJ^HPO.- 

 2CrP0, possessing a green color and retain- 

 ing three molecules of water at 98° C. 



4. The addition of disodium phosphate to 

 an acetic acid solution of chromium chloride, 

 precipitates a double phosphate of the formula 

 Na^HPO, -aOrPO,, which retains five mole- 

 cules of water at 98°, and which is consider- 

 ably lighter in color than the corresponding 

 ammonium salt. 



Reply to Criticisms of Dry Lead Defecation 

 in Raw Sugar Analysis: W. D. Horne. 

 In clarifying sugar samples for polarization 

 it is customary to add a solution of lead 

 subacetate to remove coloring matter. As the 

 precipitate formed occupies some space within 

 the 100 c.c. to which the solution must be 

 made up before filtering and observing in the 

 polariscope a corresponding concentration of 

 the solution of sugar ensues, causing the 

 polarization to be too high, in proportion to 

 the volume of the precipitate. To obviate this 

 error W. D. Home proposed that the solution 

 of sugar be made up to the full 100 c.c. before 

 clarifying and then that a powder of anhy- 

 drous lead subacetate be carefully added. 

 This has the effect of causing the precipitate 

 formed to lie outside of the 100 c.c, the acetic 

 acid radicle taking up the space within the 

 solution formerly occupied by the organic 

 radicles now combined with the lead. H. and 

 L. PeUet attacked this method, claiming that 

 1° the lead precipitate absorbed from solu- 

 tion enough sugar to counterbalance the 

 error due to the volume of the precipitate, and 

 2° the addition of anhydrous lead subacetate 

 dilutes the solution enough to account for the 

 differences of polarization by the ordinary 

 method and by Home's dry lead defecation. 

 In this paper Horne shows that the precip- 



