THE AL UNI JO URN A L. 



229 



excess of platinum chloride. When the 

 alkaloid is in excess the proportion of 

 metal is only 16.46 per cent. 



Whilst the inquiry has been in prog- 

 ress, Mr. Cripps also carried out a few 

 experiments bearing upon the chemistry 

 and pharmacy of ipecacuanha: 



1. Arndt having published further 

 information with regard to his volatile 

 alkaloids, which he states to be identical 

 with choline, the author repeated his pro- 

 cess, using 100 grammes of the root, but 

 has been unable to defect any trace of 

 volatile alkaloid. 



2. Arndt has described a process for 

 the determination of ipecacuanha alka- 

 loids, based upon his researches on the 

 volatile alkaloid. He treats the powder- 

 ed root with sodium carbonate and ferric 

 chloride in presence of boiling methyl 

 alcohol (60 percent.), under an inverted 

 condenser, filters, and evaporates the 

 filtrate, whereb)' choline is dissipated; 

 he then takes up the residue with dilute 

 ammonia, and agitates with chloroform 

 to extract the alkaloid. This solution is 

 separated, agitated with dilute acid, and 

 the aqueous liquid finally titrated with 

 Mayer's solution. The author confirms 

 Arndt's statement that this process 

 yields lower results, having obtained 1.32 

 per cent, of alkaloids from a sample 

 which yielded to Lyons' process 2.45 per 

 cent. The working of Arndt's process 

 is ver3' tedious. 



The following figures are examples of 

 the yields by various processes: 



Alkaloid Alkaloid 

 titrated. titrated. 



Process. Alkaloid weighed, eq. 254. eq. 248 



Lyons ( 1 )__. 39.3 m.g. 39.5 m.g. 38.6 m.g. 



" (2) 43.0 " 420 " 41.0 " 



Keller (1 )__. 102.0 " 87.0 '• 85.0 " 



(2)___iii.o " 95.9 " 93.6 ' 



Acetic ether. 31. 1 " 31.0 " 30.3 " 



A weaker acid yields a better result. 



Thus two portions of the root, each 



weighing half a pound, were exhausted, 



the one by the official process, the other 



by percolation with dilute acetic acid, 1 

 to 20. The extract from the latter was 

 more readily dried and pulverized than 

 that of the former. They weighed B. P. 

 900 grains; weak acid extract, 803 grains; 

 alkaloids, 8.24 per cent., and 10.6 per 

 cent, corresponding to a total of 73 

 grains and 85 grains, respectively. — 

 Paper 7 ead be/ ore the British Pharmaceu- 

 tical Conference. 



NOTE ON THE DETECTION OF FORH- 

 ALIN.* 



By H. DROOP RICHMOND and L. KIDGEIX BOSE- 

 LEY. 



A solution of formaldehyde, called formalin, 

 having come into use as a food preservative, it 

 becomes of importance to be able to detect and 

 estimate it. The literature of formaldehyde is 

 very voluminous, and numerous tests for it have 

 been proposed. As is well known, aldehydes 

 reduce Fehling's solution and ammoniacal silver 

 nitrate, and give Schiff's reaction. These reac- 

 tions, however, are by no means characteristic 

 of the aldehyde. Legler's method for the 

 estimation of formaldehyde by titration with 

 ammonia is to a certain extent characteristic of 

 formaldehyde, but it is not applicable to dilute 

 solutions. Legler states that three molecules 

 ammonia are equal to four of formaldehyde, 

 while Losekan maintains that three molecules 

 are equal to six of formaldehyde. This dis- 

 crepancy is explained by Eschweiler, who 

 shows that with methylorange, cochineal, 

 tropoeolin and congo-red, six molecules are in- 

 dicated, while with litmus and phenolphthalein 

 only four. This is due to the acid reaction of 

 the hexamethylene-tetramine formed. 



Plochl states that when a neutral solution of 

 formaldehyde is mixed with ammonium chlor- 

 ide it becomes acid ; on heating C0 2 is evolved, 

 and trimethylamiue is formed. Kleeberg 

 shows that formaldehyde combines with phenols 

 in the presence of HC1, but he did not succeed 

 in purifying the compounds formed. Pulver- 

 macher, in a series of papers, describes many 

 condensation products with substituted am- 

 monia, and also shows that the very insoluble 

 formalazine is produced by mixing formaldehyde 

 and hydrazine hydrate ; this yields a platino- 

 chloride (C a H 4 N,) e H a PtCl 6 . 



*Reprinted from the Analyst. 



