242 REPORTS ON THE STATE OF SCIENCE.— 1920. 



ultra-violet region beyond that reached with a quartz spectrograph working 

 in air Lastly, the incipient or very shallow absorption band shown by ethyl 

 «-ethoxycrotonate in the presence of acid is due to the fact that _ relatively 

 few molecules are brought into a less condensed' phase by the action of the 



It must be clearly understood that the statement that a compound only 

 shows general absorption is very misleading, because it only means that no 

 absorption band is exhibited by that compound between the spectral limits^ of 

 7000 and 2100 Angstroms. Such substances will certainly be found to exhibit 

 selective absorption when investigations are made in the very extreme ultra- 

 violet. There is a very fertile field of research in this direction by the use 

 of a vacuum spectrograph with a fluorite prism or a grating, in order to 

 obviate the absorption due to air and quartz. Some preliminary investigations 

 have already been made by Stark, who found evidences of selective absorption 

 in this region by some of the so-called diactinic substances. 



Again, the explanation of the results recorded in the examination of the 

 aromatic 'aminoaldehydes and aminoketones is very simple. These compounds 

 show one absorption band in alcoholic solution, a second in the presence of a 

 trace of acid, and a third in the presence of a great mass of acid, the frequency 

 of the second being the smallest and that of the third being the greatest. The 

 molecules exist in three different phases under the three conditions. Similarly 

 the variety of absorption bands which Hantzsch found certain substances to 

 exhibit in different solvents is due to a variety of phases of the same molecule. 

 Thus diphenylvioluric acid can be brought into several different phases by 

 alkali according to the chemical strength of LiOH, NaOH, KOH, RbOH, 

 CsOH. Further, a considerable variety of phases of trinitrobenzene, picric acid 

 and its ether trinitroanisole, can be produced by the use of solvents of different 

 basicity, such as water, alcohol, pyridine, piperidine, dimethylaniline, and 

 alcoholic f odium ethoxide.^* The case of trinitrobenzene and also of trinitro- 

 toluene is interesting, for it is possible with these two compounds not only 

 to obtain them in highly coloured molecular phases by solution in basic 

 solvents, but also to prepare these phases in the pure state. The coloured 

 liquid phase of trinitrotoluene is well known to those engaged in the manu 

 facture of this compound. The corresponding phase of trinitrobenzene can be 

 obtained by dissolving the compound in piperidine. On pouring this solution 

 into excess of hydrochloric acid the trinitrobenzene is precipitated as a red 

 solid, and after drying the colourless form may be dissolved in ether or benzene, 

 leaving the red form. A solution of this in alcohol shows the same absorption 

 band as does the piperidine solution of trinitrobenzene. 



There indeed is little doubt that the existence of a compound in two or more 

 forms as is frequently the case in organic chemistry, means the isolation of two 

 or more different molecular phases of the same compound. One of the most 

 interesting cases of the preparation of a molecular phase less condensed than 

 the ordinary phase is the so-called act-ethers of the nitrophenols.^' These com- 

 pounds on the basis of the quinonoid theory were considered to have the 

 quinonoid formula typified by 



= 



but the only evidence on which this formula was based was the instability of the 

 compounds and their visible colour. They are extraordinarily unstable and 

 change at once under the influence of certain solvents into the normal forms. 

 In the light of present-day knowledge there is not the slightest doubt that they 

 are simoly less condensed phases of the , nitrophenol ethers, having the usually, 

 acceipted formulae. 



- Reference was made previously to the quinonoid explanation of the highly 

 coloured hydrochloride of dimethylaminoazobenzene. This again is another 

 example of the conversion of a weak base into a less condensed phase by. the 

 addition of acid such as occurs with the aminoaldehydes and aminoketones. It 

 -was pointed out above that the quinonoid explanation fails because ft similar 



