TRANSACTIONS OF SECTION B. 611 



repeating the preparation in exactly the same manner, with the exception that 

 the water-bath was replaced by a sand-tray, I obtained Widman's results. 



These three modifications of benzoylphenylsemicarbazide are not only different 

 iu regard to their melting-points and solubilities in various solvents, but they can 

 be distinguished from one another by their appearance under the microscope. 



That one of them is not simply a mixture of the other two is shown by the 

 fact that when any two are ground together in a mortar the melting-point of the 

 mixture is no longer sharp, and the two modifications can be separated again by 

 the use of a suitable solvent. A mixture of equal parts of the modifications 

 melting at 205-L'06° and 211-212° began to soften at 204° and finished melting at 

 209° ; the whole of the lower melting modification could be removed by washing 

 with cold ethylic acetate. 



A mixture of equal parts of the modifications melting at 202-203° and 

 211-212° began to shrink at about 200° and was not entirely melted until 208°. 

 The lower melting substance could be removed by washing with warm benzene. 



The property of existing in three such modifications is not confined to benzoyl- 

 phenylsemicarbazide. I have obtained by similar means, that is, the action of 

 various solvents atdifi'erent temperatures, three modificationsof phenylsemicarbazide, 

 the ordinary form melting at 172°, which in its properties corresponds with the 

 Michaelis-Schmidt benzoyl derivative, and two lower melting forms melting at 

 164° and 151°, of which the one which melts at 151° corresponds in its insolubility 

 and its stability in benzene with Widman's benzoylphenylsemicarbazide. 



o-Tolylsemicarbazide, p-tolylsemicarbazide, and benzoyl-p-tolylsemicarb- 

 azide have each been found capable of existence in three modifications, and 

 from indications obtained with other semicarbazides it seems probable this would 

 be the case with all those of the general form R.R'.NH.NH.CO.NHo where 

 R is a benzenoid radical and R' is hydrogen or benzoyl. 



As to the relation of these modifications to each other, the semicarbazides in 

 question may be trimorphous, or we may have to deal with some form of stereo- 

 isomerism. Unfortunately, although we have tried a number of reactions, we 

 have not been able as yet to find one which showed a difi^erence in the chemical 

 behaviour of the three modifications, and we might tberefore accept the theory of 

 trimorphism were it not for two facts. 



The first is that it is possible to recrystallise two and sometimes all three 

 modifications from the same solvent, and even in presence of one another, without 

 conversion of one into another. That seems to me to dispose of trimorphism. 



The second fact is tbat with certain semicarbazides. diphenvlsemicarbazide 

 (C,H,),N.H.CO.NHj, hydrazodicarbamide :S^H,,.CO.NH.NH.CO.NH,, and benzal- 

 semicarbazone C^H^.CH.N.NH.CO.NHj, I have been unable to obtain even a 

 suggestion of a second form. 



That seems to point to the property of existing in these different forms being 

 dependent on the nature of the groups in position 1. 



NH„.NH.CO.NHa. 



(1) (2) (3) (4) 



At present, however, we have not sufficient evidence on which to base any 

 theoretical explanation. 



8. Report on Isomeric Naphthalene Derivatives. — See Reports, p. 152. 



9. Report on Isomorphous Derivatives of Benzene. — See Reports, p. 78. 



