230 



THE ALUMNI JOURNAL. 



Trillat gives the following tests for formalde- 

 hyde : The solution is to be mixed with dime- 

 thylaniline acidified with sulphuric acid and 

 agitated. After heating for 30 minutes on the 

 water bath, it is made alkaline, boiled until the 

 smell of dimethylaniline has disappeared, then 

 filtered. If the filter paper be moistened with 

 acetic acid and powdered lead oxide be sprinkled 

 on it, a blue color, due to the formation of 

 tetramethyl-diamido-diphenylmethane is pro- 

 duced if formaldehyde is present. Or the 

 formaldehyde solution may be mixed with a 

 solution of aniline (3 Gm. to 1 liter), when a 

 white precipitate of anhydro-formaldehyde 

 aniline appears which may be weighed. A pre- 

 cipitate is also given by acetaldehyde. Trillat 

 says that, as formaldehyde easily forms conden- 

 sation products, it is not always detected in food 

 after a lapse of time. 



Three years ago one of us worked with 

 formaldehyde as a preservative for milk, and 

 used as a method for its detection the reduction 

 of Fehling's solution, or of ammoniacal silver 

 nitrate solution. Quite recently Thomson 

 {Chem. News, lxxi., 247) has proposed the use 

 of the latter, and, although he modifies the test 

 by working in the cold, he does not succeed in 

 obtaining a reaction which is characteristic of 

 formaldehyde. 



Schiff's reagent has been used as a test for 

 formaldehyde. It is, however, very unsatis- 

 factory unless care be taken, for if an excess of 

 sulphurous acid is used, no reaction is obtained 

 with traces of formaldehyde, and any alkali 

 combined with an acid weaker than S0 2 also 

 gives a red coloration. The red coloration ap- 

 pears on warming Schiff's reagent, on blowing 

 air through it, or even on placing it in an un- 

 corked bottle, so that unless great precautions 

 are taken the test is unreliable. Still, it is use- 

 ful as a confirmatory test. In applying it as 

 such to milk, we precipitate the casein with a 

 little sulphuric acid, filter and then add a little 

 Schiff's reagent to the filtrate ; any red color 

 which may appear roughly indicates the amount 

 of formaldehyde present. 



Another test which we believe to be well 

 known, though it has not actually appeared in 

 print, was pointed out by Hehner. It is the 

 formation of a blue color when milk, formalde- 

 hyde and sulphuric acid are mixed together. 

 This was first brought to our notice by Bevan, 

 who had obtained a blue color in a Leffman- 

 Beam experiment, which he could not account 

 for. We suggested that it might be due to 

 formaldehyde, but we were unable to obtain 



the reaction with other milks, owing to our hav- 

 ing used an excess of formaldehyde. We find 

 that when formaldehyde is in large quantity, 

 say 0.5 per cent., no blue color is obtainable. 



We have since found that the above reaction 

 is due to the albuminoids of milk. We have 

 also obtained it from egg albumen and peptone, 

 but not from gelatin. To obtain the reaction 

 it is only necessary to add sulphuric acid (95 

 per cent. H 2 S0 4 gives the best results) to the 

 milk, when a blue ring is formed at the junc- 

 tion of the two fluids. The food suspected may 

 be distilled and the formaldehyde obtained in 

 plain aqueous solution ; but we prefer the use of 

 peptone for testing, as the blue color is not then 

 obscured by the charring of the organic matter 

 by the acid. 



Bearing in mind Pulvermacher's researches, 

 we have found a reaction between formalin and 

 diphenylamine. A solution of diphenylamine 

 in water is made, just sufficient sulphuric acid 

 being added as will effect solution. The liquid 

 to be tested (or the distillate), is added to this 

 solution and boiled. In the presence of formal- 

 dehyde a white flocculent precipitate is deposit- 

 ed, which is often colored green if the acid used 

 contained nitrates. 



We find it most convenient to distill into the 

 diphenylamine solution and then boil. This 

 simple test we believe to be characteristic of 

 formaldehyde. 



We are engaged in determining the composi- 

 tion of the precipitate, and in working out the 

 quantitative estimation of formaldehyde in this 

 manner. We are able to confirm Trillat's ob- 

 servation that after a certain time formaldehyde 

 cannot be detected. We can obtain the reac- 

 tion in milk which has not curdled . We think 

 from the list of methods enumerated that there 

 is not the slightest difficulty in definitely prov- 

 ing the presence of formaldehyde in foods when 

 present. Hehner's reaction, confirmed by the 

 diphenylamine test, Schiff s test, and those pro- 

 posed by Trillat, Pulvermacher and Plochl, 

 should be amply sufficient. 



Caffearine. — Paladino has isolated from cof- 

 fee beans a new alkaloid, crystallizing in 

 needles, which he proposes to name caffearine. 

 It is precipitated by potassio-bismuthic iodide 

 after the removal of theine by shaking with 

 chloroform. From the precipitate the alkaloid 

 can be regenerated. — The Druggists' Circular 

 and Chem. Gazette. 



