610 REPORT — 1901. 



in benzene, with benzoic chloride, but found the melting-point of the product to 

 be 210-211°. 



C,H,.NH.NH.CO.NH, + C,H,.C0C1 = C,;H,.N(00C,H5).NH.CO.NH, + HCl. 



Widman then repeated the preparation, making use of the Michaelis-Schmidt 

 method, and again obtained a product melting at 210-211°. From this it might 

 have been inferred that the lower melting-point obtained by Michaelis and 

 Schmidt was probably due to insufficient purification. 



Towards the end of 1896 I had occasion to prepare benzoylphenyUemicarb- 

 azlde, and made use of Widman's method. The product obtained melted at 

 202-203°, and the melting-point was not altered by repeated recrystallisations. I 

 had thus obtained the Michaelis-Schmidt product by means of Widman's method 

 of preparation, and at the same time confirmed the Michaelis-Schmidt melting- 

 point. These results having been published in the ' Transactions of the Chemical 

 Society,' Dr. Widman kindly sent to me a sample of his benzoylphenylsemi- 

 carbazide, which I -found by my thermometer to melt at 210-212°, so that the 

 thermometers were not to blame for the discrepancy. 



A comparison of the two substances showed that the one which melted at 

 202-203° was distinctly more soluble than the higher melting one in the 

 ordinary solvents, such as benzene, ether, alcohol, or water. 



The first interesting observation was that on boiling a small quantity of 

 Widman's substance with water: it went into solution very slowly, and on 

 coolinf separated in crystals, which melted at about 200°, and, after recrystallisa- 

 tion trom dilute alcohoX at 202-203°. 



The Widman form of benzoylphenylsemicarbazide had been transformed into 

 the Michaelis-Schmidt form. 



The reverse transformation was not so easy, partly because it takes place, even 

 under the best conditions, very slowly ; partly because it leads under varying 

 conditions, not only to Widman's benzoylphenylsemicarbazide, but also to a third 

 modification. 



This third form of benzoylphenylsemicarbazide melts at 205-206°. It is 

 produced from the lowest melting form by boiling with benzene in a reflux 

 apparatus on the water-bath. The change took place with 2g. substance in 

 about fifteen hours, and further similar treatment was without effect. 



This benzoylphenylsemicarbazide, which melts at 205-206°, can be re- 

 crystallised unaltered from benzene, ethylic acetate, acetone, or alcohol. Boiling 

 water and, more slowly, dilute alcohol convert it into the lowest melting form. 



If the Micliaelis-Schmidt product be boiled with benzene in a reflux apparatus 

 on a sand-tray, instead of on a water-bath, the result is diflerent. The change, 

 here again, is slow. After some fifteen hours' boiling the product softens at 

 about 205°, and is completely melted at 210°. Further boiling is without eflect. 

 This product can be separated, by washing with cold ethyhc acetate, into 

 Widman's benzoylphenylsemicarbazide, which remains undissolved, and the 

 third modification which dissolves in the ester. 



By this method of boiling with benzene on a sand-tray the form melting at 

 205-206° can be converted partially, but never completely, into Widman's 

 modification. 



The process of lowering the melting-point from 211-212° to 205-206° is best 

 carried out by boiling with acetone. 



The difierence in effect produced by exchanging the water-bath for a sand-tray 

 in boiling with benzene suggested what is probably the reason why, although 

 makino- use of apparently exactly the same method as Dr. Widman in preparing 

 the benzoylphenylsemicarbazide, I obtained, not his product, but that of Michaelis 

 and Schmidt. 



In the preparation phenylsemicarbazide and benzoic chloride are boiled in 

 benzene in a reflux apparatus. I had carried out the boiling on a water-bath, and 

 obtained benzoylphenylsemicarbazide of melting-point 202-203°, mixed, as was 

 found later, with a small quantity of the form which melts at 205-206". On 



