602 ON PLANT GROWTH HORMONES 
The experiment was repeated twice with the acid, and once with the lactone ; 
we were thus able to characterise the degradation product by means of 
analyses, titration, and preparation of the p-Phenyl phenacyl ester as a 
dicarboxylic acid of the formula C,;H5,0,4. 
Of course one is inclined to assume that the auxin molecule was split 
at the double bond, and that one of the carboxyl groups of the dicarboxylic 
acid was identical with that one already present as such in the auxin molecule. 
This assumption, however, at once creates a difficulty: since our C,3-acid 
does not contain the hydroxyl group which in auxin occurs in the $-position 
relative to the carboxyl, this group must have been present in the C; residue 
removed by oxidation. If we do not wish to assume a lactone ring with 
more than six members, we arrive at the partial formula of Fig. 4, according 
to which the C,;-acid should be a substituted malonic acid. The acid, 
however, could be heated 100° above its melting point, without losing carbon 
dioxide. We assume therefore that both carboxyls of the C,;-acid have been 
Cra th 0, + CsthO, 
(auxin) (arcarfoxylic acta) 
— C7 
AH 
R- Git: v NC 
COOH- 
This formula would ead 
tod malonic acid: 
v2 CA COOH 
Double Bond betwe COOtt 
Cand Cy ssainael wou Gsad toan 
austin hes no 
enolic hiydrosgt ? 
Fic. 4. 
newly formed by the oxidation of auxin, whilst the ‘ auxin carboxyl’ 
originally present was removed with the C;-residue. 
In order to test this hypothesis further, we oxidised dihydro auxin in 
glacial acetic acid solution with chromium trioxide. So far we could only 
do one experiment, with 22 mg. of substance. Fortunately we obtained 
besides oxalic acid a neutral product which could be converted into a crys- 
talline p-nitrophenyl hydrazone, the micro-analyses of which indicate the 
formula C,;H;,O for the oxidation product. Since it gives no aldehyde 
reactions it must be a ketone. In this oxidation also the oxygen atoms 
originally present in the auxin molecule have disappeared with the C; 
residue ; evidently the cyclic ketone corresponding to the C,3-acid has been 
formed. 
The simplest explanation of the results of the degradation leads to the 
following working hypothesis (Fig. 5) : 
(1) The three hydroxyl groups are not distributed over the whole mole- 
cule, but are localised in the region of five carbon atoms ; one of these five 
carbon atoms belongs to the carboxyl group. 
(2) In the C;-residue there is probably a hydroxyl group in the 8-position 
with respect to the carboxyl, whilst the y-position is free of hydroxyl. The 
two other hydroxyls would then have to be in positions « and 8. : 
(3) The ring of auxin is not terminal ; it probably contains the double 
