76 PROCEEDINGS OF THE AMERICAN ACADEMY 



From the work described above, it appears that 20 grs. of ortho 

 brombenzylbroraide yielded, — 



Total organic product 10.850 grs. 



Anthracene 1.409 " 



9.441 " 

 Phenanthrene 0.190 " 



9.251 " 

 Dibromdibenzyl, or its isomeres, calculated from 



3.17 grs. of unremoved bromine 6.738 " 



Undetermined, containing phenanthrene, dibenzyl, 



benzyltoluol, and ditolyl 2.513 " 



Summary of Results. 



Orthobrombenzylbromide when treated with sodium yields anthra- 

 cene (identified by its melting-point 212°, the formation of anthrachi- 

 none melting-point 273°, and of alizarine; 19.78% of the theoretical 

 yield) ; and phenanthrene (identified by the melting-point of its chi- 

 none 198°, and by Laubenheimer's test; 2.66% of the theoretical 

 yield). Certainly part, and probably all, of the anthracene occurs in 

 the free state ; no anthracenedihydride could be found. Our experi- 

 ments leave us in doubt as to the way in which the phenanthrene occurs. 

 Dibenzyl was also obtained, an oil possibly benzyltoluol or ditolyl, 

 and a viscous substance containing bromine. 



By this synthesis of anthracene from orthobrombenzylbromide, it is 

 proved that the tico connecting carbon atoms are attached to each ring 

 in the ortho position, and thus the last doubt about the constitution of 

 anthracene is removed. 



The synthesis of phenanthrene, while it confirms the ortho attach- 

 ment of the two additional atoms of carbon, proves nothing either for 

 or against the formula recently proposed by Merz and Weith. 



The publication of the preliminary notice of this work, already 

 mentioned, drew out an article from Von Peclmiann,* describing work 

 done at about the same time as ours, in which he applies to anthra- 

 cene the ingenious method contrived by Graebe for determining the 



* Ber. d. ch. C, '79, p. 2121. 



