78 
hands have turnished important clues to the structural 
formule of terpenes and other compounds. 
As additive properties become constitutive, so the 
value of a knowledge of the physical properties of a 
substance will tend to increase, but there is little 
ground for hope that the problem of the constitution 
of benzene will be solved from the physical side. The 
controversy which has arisen between Hantzsch and 
Auwers regarding the physical properties of cyclo- 
tatetraene in relation to its chemical structure is a 
case in point;*® the absence of optical exaltation in 
this hydrocarbon is wholly unexpected, but, on the 
other hand, the type of compound is entirely new. 
With benzene also the distribution of valency within 
the molecule differs from that in any known com- 
pound; our knowledge of it, admittedly far from 
complete, has been gained from the chemical side, and 
is summarised in the various structural formule; but 
the limitations of the physical method of attack can 
be traced from Thomsen’s endeavour to determine its 
structure from thermochemical data** to the more 
recent invention of isorropesis. And, despite the 
evidence obtained from refractivities, we may not un- 
reasonably demur to the suggestion that derivatives 
of benzene, which by their behaviour towards sub- 
stituting agents show themselves to be wide apart in 
chemical properties, such as nitrobenzene and aniline 
on one hand or chlorobenzene and phenol on the other, 
should respectively be classified together.**. Un- 
doubtedly, most useful information is obtained from 
a comparison of the physical properties of two related 
substances, the exact constitution of one of which is 
uncertain, but that of the other known. Therefore, 
bearing in mind the great development that has taken 
place recently in the correlation of physical properties 
with chemical constitution by methods based on re- 
fraction and absorption, every chemist will welcome 
the entry of Dr. Lowry into that field of research on 
the relation between magnetic rotation and structure, 
which for all time will be associated with Sir William 
Perkin’s name. 
Substitution in the Benzene Series. 
Turning now to a discussion of the problem of 
substitution in cyclic compounds, one important factor 
must not be overlooked; the even distribution of the 
residual affinity of the benzene molecule is disturbed 
by the introduction of a substituent. The study of 
substitution in benzene derivatives indicates that, as 
a consequence of this disturbance, a directing influ- 
ence comes into play which, when the substituent is 
changed, may vary in the effect it exercises on the 
course of substitution. 
Arising probably from this even distribution of 
valency, it is characteristic of benzene to furnish addi- 
tive compounds in which six atoms of hydrogen or a 
halogen, but not two or four, become attached sym- 
metrically to the molecule; substitution, however, 
occurs when a catalyser is present, such as the 
aluminium-mercury couple for halogenation, or sul- 
phuric acid for nitration or sulphonation, leading 
initially to the production of mono-substituted deriva- 
tives. Whether the catalyser by association with the 
benzene molecule*® limits this additive capacity, or 
whether its function is to promote the elimination of 
the halogen acid or water respectively,*® is still a 
subject of discussion, but in the absence of a reaction 
36 A. Hantzsch, Ber., 19:2, xlv, 563; K. Auwers, zd/d., 971. 
37 Cf H. E. Armstrong, Phil. Mag., 1887 [v.] xxiii, 73; J. W. Brithl, 
J. pract. Chem., 1887 [ii], xxxv, 181, 209. 
38 Cf J. W. Rriihl, Zeit. Physikal. Chem., 1804, xvi, 220; Smiles, ‘‘ Re- 
lations between Chemical Constitution and Physical Properties” (Longmans, 
1910), p. 299. . te 
39°B, N. Menschutkin, Abstr. Chem. Soe , 1912, cil, i, 98-100. 
40 Cf H. E. Armstrong, Trans. Chem. Soc., 1887, li, 263. 
NO. 2290, VOL. 92] 
NATURE . 
[SEPTEMBER 18, 1913 
of additive type it is not easy to account for facts 
such as the production of a certain amount of tri- 
nitrophenol when benzene is nitrated in the absence | 
of sulphuric acid. 
wa 
Ne 
—H,O 
H.OH —HNO, 
= 28 28 
a é OH# 
| | NO H.NO Ap f 
ace NU \Z 
(in presence of 
sulphuric acid) 
(in absence of 
sulphuric acid) . 
The much-debated questions still remain: Why and 
by what mechanism, when a second or third substi- 
tuent is introduced into the molecule, is the orientation 
of the isomeric products determined by the radical or 
radicals already present? For disubstitution, the 
ortho-para- and the meta-laws have been deduced, and 
the radicals which respectively promote mainly ortho- 
para-substitution on one hand, and meta-substitution 
on the other, have been catalogued.*! But these laws 
take account only of the orientation of the chief pro- 
duct or products, whereas all three derivatives, ortho, 
meta, and para, have been detected in most of the 
reactions studied, and their relative proportion in 
many cases is known to depend on the conditions, 
being affected by such influences as variation in tem- 
perature or in the medium employed.*? Nitration of 
acetanilide, for example, furnishes a mixture of ortho- 
and para-nitracetanilide, but of aniline in the presence 
of much sulphuric acid yields chiefly meta- 
nitraniline.*® And, to illustrate the inadequacy of the 
meta-law, the fact that sulphonation of benzene- 
sulphonic acid with concentrated sulphuric acid at 
230°-240° furnishes an equilibrium mixture of the 
meta~ and para-disulphonic acids in the proportion of 
2:1 may be quoted.** 3 
In the exploration of this field many workers have 
participated, but the results, recorded almost as often 
in patent specifications as in journals, are seldom 
quantitative, so great is the difficulty at times in 
isolating the minor product or products of the change. 
Recently, however, by a most ingenious use of melt- 
ing-point curves and density determinations, Holleman 
and his collaborators have carried out an exhaustive 
series of substitutions with small quantities of mate- 
rial and under known conditions;*® yet after a 
survey of the whole field the conclusions reached 
are :— 
(1) That uncertainty cannot be removed until some 
basis exists for different reactions to be carried out 
under comparable conditions.*® 
(2) That even if the relative amounts of the iso- 
merides formed when a radical C is introduced into 
each of the mono-substitution derivatives C,H,;.A and 
C.H;.B be known, it is not possible to calculate the 
proportion in which the isomerides C,H,.ABC will be 
produced when the radical C is substituted in the 
compound C,H,.AB. 
Although the validity of the ortho-para and of the 
meta-laws may be impeached, they serve as a first 
approximation, and many theories have been pro- 
pounded to account for them. Armstrong has sug- 
gested that in ortho-para-substitution the additive com- 
pound is formed by association of the addendum with 
the carbon atom carrying the radical already substi- 
4. The phenol by nitration forming the trinitro- derivative (picric acid), 
Armstrong and Rossiter, also Groves. Proc. Chem. Soc., 1891, vii, 89. 
42 Cf. Noelting, Ber., 1876, ix, 1797: Armstrong, Trans. Chem. Soc., 
1887, li, 258 ; Crum Brown and Gibson, 7éd., 1892, Ixi, 367. 
43 Hiibner, Anmaden, 1881, ccviii, 2¢9. 
4 J. J. Polak, Rec. trav. chim, 1910, [ii.] xiv, 416; R. Behrend and M. 
Mertelsmann, Amalen, 1911, ccclxxviii, 352. 
4 A. F. Holleman, ‘‘ Die direkte Einfithrung von Substituenten in den 
Benzolkern” (Leipzig, Viet and Co., I9I0), P- 215- 
46 For example, nitration is effected chiefly at low temperatures, but sul- 
phonation of mono- substituted benzenes at temperatures higher than the 
ordinary, which if employed in nitration would lead to mixed products. 
