10 BULLETIN 227, U. S. DEPARTMENT OF AGRICULTURE. 
acids being only slightly toxic. With the organic acids, in which 
ionic dissociation is usually less complete, he also found that the 
relative importance of the H ions in general varied with the per- 
centage of dissociation. If the dissociation was relatively slight, the 
nonionized molecule itself exerted the predominating influence. In 
general, the anions of organic acids were found to possess relatively 
slight toxic properties, oftentimes so slight as to be almost negligible, 
and, since both the sodium ions and the anions were usually but 
weakly toxic, it followed, as a rule, that sodium salts showed but 0.5 
to 3 per cent of the toxic value of the corresponding acids. Carboxyl 
hydrogen proved much more toxic than hydroxyl hydrogen. Since 
in the phenols this latter form of combination occurs, and since these 
substances do not ionize, the toxicity here must be referred entirely 
to the undissociated molecule. 
In order to throw further light on the behavior of phenols and their 
derivatives True and Hunkel (30) extended their investigations on 
Lupinus albus to this group. The results bear out their earlier con- 
clusions that electrolytic dissociation of phenylic bodies plays but a 
very subordinate role in determining their toxicity. However, in a 
few instances, such as with picric and salicylic acids, the cresols, and 
the mononitrophenols, electrolytic dissociation is said to exert a pro- 
nounced influence. Some phenols also, like pyrocatechol and hydro- 
quinone, which are comparatively unstable, may quickly change to 
constituents even more fatal than H ions. Certain radicals seemed 
also to have specific properties when introduced into the molecule. 
For instance, the number of hydroxyl groups appeared to have little 
influence, while the introduction of the methyl group into the benzene 
nucleus increased the toxicity to a considerable but variable degree, 
as shown by the cresols and less plainly by orcinol; however, replacing 
the H of a hydroxyl group by a CH 3 group had little effect. The 
introduction of the isopropyl group into the cresols further increased 
their toxicity. The presence of one or more nitro groups likewise 
increased toxicity to a great degree, but the number of these groups 
seemed to make little difference. 
Similarly, in the case of certain organic compounds (cf. 7, pp. 351- 
352), Ehrlich and Bechhold have shown that the introduction of 
halogen and alkyl groups into the benzol ring increases the toxicity 
of phenols to diphtheria bacteria, two molecules of pentabrom- 
phenol being about equal to 40 molecules of trichlorphenol and 100 
molecules of phenol. On the other hand, the introduction of the car- 
boxyl group was said to lessen toxicity. 
Likewise, Falck (7, pp. 355, 357) states that nitrophenols and dini- 
trophenols are considerably more toxic than phenol and more so when 
the nitro groups are in the ortho position than when in the meta or 
para position. The most effective of 19 nitrophenols which he tested 
