422 PROCEEDINGS OP THE AMERICAN ACADEMY. 



low as 40°. After we had abandoned for the present the attempt to- 

 analyze the green salt, we succeeded in throwing some light on its 

 composition by determining the proportion of quinone to sodic ethylate 

 necessary to form it, as we found that each molecule of quinone takes 

 up one molecule of sodic ethylate, but it must be left to future experi- 

 ments to decide whether the product is really a hemiacetal. 



In all the work so far described in this paper only two of the atoms 

 of chlorine in chloranil have been replaced by other radicals; we 

 found, however, that the other pair of atoms of chlorine could be 

 replaced by phenoxy groups, if the dichlordiphenoxyquinone was 

 treated with sodic phenylate, or if chloranil was acted on by four 

 equivalents of sodic phenylate. The tetraphenoxyquinone thus formed 

 melted at 229°-230°, and offered a rather striking resistance to the 

 action of reducing agents, although zinc dust and glacial acetic acid 

 converted it into tetraphenoxyhydroquinone, C6(OCgH5)^OH, melting 

 point 219°-220°. Toward acids the tetraphenoxyquinone shows a 

 marked stability, but by boiling with a strong solution of sodic hydrate 

 it was converted into the diphenoxanilic acid, Cg(OCyIl5)o(OH)o02, 

 which melts at about 276°. Sodic methylate converts tetraphenoxy- 

 quinone into the dimethoxydiphenoxyquinone by replacing two of its 

 phenoxy by methoxy groups ; the substance melts at 171°. The 

 corresponding diethoxydiphenoxyquinone, melting at 128°, was formed, 

 instead of tetraphenoxyquinone, when chloranil was treated with sodic 

 phenylate made from phenol and sodic ethylate in alcoholic solution. 

 It is a noteworthy fact that bromanil acts differently with sodic phe- 

 nylate made in this way, giving the dibromdiphenoxyquiuone. One 

 other case was observed of the substitution of all four of the chlorine 

 atoms of chloranil. This was when the dichlorquinone dimalonic ester 

 of Stieglitz was boiled with alcohol and sodic carbonate, as it was 

 converted into diethoxyquinone dimalonic ester melting at 115°. It is 

 certainly strange that such a rather weak reagent should remove these 

 two atoms of chlorine, which in other cases have seemed very firmly 

 attached to the molecule. In all the reactions just described it is to 

 be observed that the chlorine atoms or phenoxy groups are replaced in 

 pairs, and this fact also appears in much of the work with chloranil 

 previous to ours. This replacement of the radicals two at a time can 

 probably be connected with the para position of the two atoms of 

 oxygen, which in this case serve to diminish the attraction of these 

 radicals to the benzol ring, and thus make it possible to replace them, 

 as in the case of tribromdinitrobenzol (BrNOoBrNOoBr), melting point 

 192°, where the loosening nitro groups are in the meta position, all 

 three of the bromine atoms are replaced in many reactions. 



