460 The Oxydases of Cytisus Adami .
By Frederick Keeble , Sc. D. , Professor of Botany , University College , Beading , and E. Frankland Armstrong , Ph. D. , D.Sc .
( Communicated by W. Bateson , F.K.S. Beceived May 29 , \#151 ; Bead June 20 , 1912 .
) The graft-hybrid Cytisus Adami\#151 ; that classic of botanical speculation and research\#151 ; is one of the wonders of the vegetable world .
Every year the trees , which have been propagated vegetatively from the original creation of the French gardener Adam , may be seen to bear blossoms of three kinds , buff , yellow , and purple .
The yellow flowers are identical with those of the common laburnum ( Cytisu laburnum ) , the purple resemble those of Cytisus purpureus , and the buff coloured flowers appear to combine the characteristic features of C. laburnum with those of C. purpureus .
The history of the origin of C. Adami has been told repeatedly { cf. Darwin , 1888 ) .
There is good ground for believing that it was produced by budding C. purpureus on C. laburnum .
The plants raised as a result of the operation were\#151 ; according to Darwin 's account\#151 ; distributed originally as C. , and it was not till later , when they began to exhibit their diversity of flower , that they were described as graft-hybrids .
Notwithstanding the attention that has been given to the subject , the nature of the union between the two species remained obscure for many }rears .
Becently the brilliant researches of Baur ( 1909 ) and Buder ( 1910 ) have led the former author to formulate an hypothesis which has the signal merit of presenting a precise and diagrammatic picture of the biological construction of C. Adami .
On this hypothesis , the graft-hybrid is to be regarded as a periclinal chimera ; that is to say , it is a dual organism composed externally of one species and internally of another .
The contribution of purpureus to the chimera consists of a superficial single-layered epidermis ; the rest of the body of C. Adami is built up of tissues contributed by C. laburnum .
The hypothesis , which is supported by the results of investigations into the comparative anatomy of the component species and the graft-hybrid , and also by analogy with other graft-hybrids , permits of a postulation of the modes of origin of the several kinds of flowers borne by C. Adami .
Thus , branches of the graft-hybrid in which the periclinal duality is maintained bear flowers with a purple epidermis and yellow sub-epidermal cells .
The combination results in a buff colour .
Those branches which , for some reason or other as yet unexplained , are produced solely by the sub-epidermal laburnum The Oxydases of Cytisus Adami .
461 cells bear yellow flowers of the laburnum type .
The purple flowered branches arise exclusively from epidermal cells , and are , therefore , similar to those of C. purpureus .
The investigations , the results of which are now to be described , were undertaken with two objects .
In the first place , they were designed to test the validity of Baur 's hypothesis , and , in the second place , to determine whether the results of an examination of the distribution of oxydases in G. Adami and its component species would confirm the hypothesis , to which our previous work ( Keeble and Armstrong , 1912 ) has led us , that oxydases may migrate from one part of a plant to another .
We may say at once that the results of our investigation of the distribution of oxydases in C. Adami serve to confirm the general truth of Baur 's conclusions and to support the thesis that oxydases may migrate from cell to cell .
# The methods which we employ for determining the distribution of oxydases in plant-tissues are described in our former contribution .
The results obtained by the use of benzidine are as follow- : Petals of G. purpureus treated with an alcoholic solution of benzidine and incubated at 37 ' C. for half an hour are first decolorised and then give a direct oxydase reaction , that is one which declares itself without the addition of hydrogen peroxide .
The reagent produces a grey to bluish-brown coloration of the general surface of the petals , more marked in the marginal region , and a similar but deeper coloration of the veins .
In other words , G. purpureus contains an epidermal oxydase and also a bundle oxydase .
The petals of the purple ( magenta ) flowered branches of G. Adami\#151 ; which branches are , on Baur 's hypothesis , pure G. purpureus\#151 ; give similar , albeit somewhat fainter , reactions , thus indicating that the oxydases of the purple branches of G. Adami are identical in kind and distribution with those of G. purpureus .
When treated with benzidine the typical buff-coloured flowers of G. Adami do not become completely decolorised .
This is owing to the fact that they contain , in addition to an anthocyanic epidermal pigment , a yellow pigment of the carotin type , which lies , in both G. Adami and G. laburnum , in the sub-epidermal cells .
Nevertheless , despite the fact that the carotin pigment is present , it is easy to determine , either by direct inspection or by microscopic examination of surface sections of the petals , that the buff-coloured flowers of G. Adami contain epidermal oxydase .
In this respect C. Adami differs from the yellow C. laburnum and from the labumum-like flowers of the yellow-flowered branches of G. , neither of which contains epidermal oxydase .
We conclude , therefore , that the epidermis of G. purpu/ reus contains an oxydase , that this oxydase is present 462 Prof. F. Keeble and Dr. E. F. Armstrong .
[ May 29 , also in the epidermis of G. Adami\#151 ; both in the buff-eoloured typical G. Adami flowers and in the purple flowers borne on C. Adami ; but that it is absent from the epidermis of G. laburnum and from that of the yellow flowers of C. Adami .
Although , as just indicated , the petals of and C. Adami give similar epidermal oxydase reactions they differ from one another with respect to the mode of reaction of their deeper tissues .
For whereas the petals of both G. purpureus and of purple-flowered branches of Adami give , with benzidine alone , a well-marked reaction for bundle oxydase , the vascular bundles of the buff petals of G. Adami , those of the yellow petals of the laburnum-like branches borne on the graft-hybrid and those of the petals of G. laburnum give no direct oxydase reaction .
In the three last-mentioned cases it is only after the addition of hydrogen peroxide that the bundle oxydase reaction is developed .
Whence it is to be concluded that oxydase is present in the tissues which accompany the veins of the petals of G. purpureus and purple-flowered branches of G. Adami , and that peroxydase is present in the corresponding tissues of the buff and yellow flowers of G. Adami , and in those of the common laburnum .
A closer examination of the brown bundle oxydase reaction produced by benzidine shows that it is most marked in G. purpureus , fairly marked in the purple-flowered branches of G. Adami , and that it does not occur in the buff flowers of typically Adami branches .
This is in accord with the differences in pigmentation of the three flowers .
For in the living flowers of G. purpureus lines of darker purple ( magenta ) accompany the veins of the petals .
Similar though less well-marked magenta lines run with the veins of the purple flowers of G. Adami , whereas there are no such lines of magenta colour along the veins of the buff G. Adami petals .
We conclude , therefore , that in flowers the oxydases which accompany the veins ( bundle oxydases ) co-operate with those contained in the epidermal cells to produce the pigment of those cells .
If this view be accepted , the failure of C. Adami , although its epidermis is that of G. purpureus , to produce flowers of as deep a purple as those of G. purpureus , is intelligible ; for the vascular system of G. Adami is derived from the G. laburnum component , and in that species the vascular tissues of the flower\#151 ; though they contain peroxydase\#151 ; possess no direct oxydase , and it is therefore reasonable to suppose that they cannot contribute to pigment formation in the epidermis .
If , further , we assume , and it is an assumption for which grounds are not altogether lacking , that oxydases may travel through the vascular tissues , it is possible to account for the fact that the purple pigment of the flowers of G. Adami is generally paler than that of the flowers of G. purpureus .
1912 .
] The Oxydases of Cytisus Adami .
The former receive only the oxydase of the branch , the latter have at their disposal contributions from the whole plant .
The conclusions based on the results of treating petals with benzidine are confirmed by those obtained with a-naphthol .
For although this reagent produces no definite epidermal reaction , * it gives rise to distinct bundle oxydase reactions in the flowers of G. purpureus and purple C. Adami .
On the other hand , a-naphthol gives no direct oxydase reactions , neither in the flowers of G. laburnum .
, typical G. Adami , nor in those of the yellow branches of G.Adami .
In these last-mentioned forms a-naphthol and hydrogen peroxide produce the characteristic colour-reaction , whence it is to be concluded that they contain peroxydase .
As indicated in the accompanying diagrammatic representation of the results described above , the accuracy of Baur 's representation of C. Adami as a periclinal chimera is confirmed by a study of the oxydases of the flowers The Oxydases o Cyhisus Ad\amp ; mi C. Adajn i. Purple/ lowers --------- ?
- Bi^yyiowers -------------------i Yellow/ lowers ---------of the component species and of the synthetic form .
Whether , however , the growing points of G.Adami consist always of a single-layered mantle of G.purpureus cells enveloping cells solely of the G. laburnum type must remain for the present uncertain .
It is not too much to suppose that , if the migration of oxydases is finally proved and shown to be of general ' occurrence , considerable light will be thrown on the nature of the , at present , puzzling differences in form between the flowers of G. Addmi and those of the component species , and in particular it may help to elucidate the causes of differences which exist between the flowers of G. 'purpureus and those of purpureus branches of the graft-hybrid .
The facts now to be mentioned appear to point distinctly to a migration of oxydase from cell to cell .
The most elegant part of Baur 's demonstration of the peri-chimeral nature of G. Adami has reference to small groups of cells containing red ( anthocyan ) pigment which occur in the neighbourhood of the median line on the lower parts of the standards of flowers of G. Adami* These patches of pigmented cells\#151 ; described sometimes as honey-guides\#151 ; though visible on the surface of the intact flower are situated in the * See Keeble and Armstrong , loc. cit , / epidermal direel-oxydase V C. purpureus t__________ { no epidermal oxydase C. Laburnum VOL. LXXXV.\#151 ; B. 464 Prof. F. Keeble and Dr. E. F. Armstrong .
[ May 29 \#171 ; ub-epidermal layer .
They must , therefore , as Baur points out , he formed by the G. laburnum component of the chimera .
Examination of the flowers of " .
laburnum shows the presence of these red-brown or chocolate patches on the standards and hence they supply Baur with a convincing argument that the sub-epidermal tissues of G. Adami are derived from the laburnum ( G. laburnum ) .
Now microscopic examination of transverse sections across the patches of pigment cells on the standard of a flower of C. Adami shows at once that the cells , of purpureas origin , which overlie the sub-epidermal pigmented cells of laburnum origin , are always more deeply pigmented than are the neighbouring ^epidermal cells , which have no anthocyan-containing cells immediately ^beneath them .
In other words the deeper pigmentation of the epidermal purpureas cells coincides with the distribution of soluble ( anthocyan ) pigment in the underlying and contiguous laburnum cells .
In other parts of the petals , the laburnum cells contain only yellow carotin and there the purple colour of the epidermal cells remains faint .
Since , by the nature of the case , pigment-containing cells are impermeable to their soluble pigments , it seems very probable that the heightened colour of the epidermal cells which are contiguous with the red patches of sub-epidermal cells is due to a migration of oxydase from the sub-epidermal , laburnum cells , to the epidermal , purpureus cells .
It might perhaps be urged that the paleness of the epidermal cells which overlie the yellow tissues is due to some inhibition of oxydase on the part of the latter .
This , however , does not appear to be the case .
For as we have shown ( loc. cit. ) it is possible in other flowers ( )to destroy the inhibitor of oxydase which exists in dominant white flowers by treating the petals with hydrogen cyanide .
This reagent , however , when applied to flowers of G. Adami does not exert any influence on the strength of the oxydase reactions subsequently obtained .
Hence we conclude that the yellow cells do not exercise an inhibitory effect on the pigment-forming action of the oxydase in the epidermal cells .
Finally , we would point out that the hypothesis of oxydase migration , not only from sub-epidermal to epidermal cells , but also from the vascular tissues to the epidermis , is in conformity with what is known with respect to the variable behaviour of many florists ' flowers .
Thus , in certain seasons wallflowers , and in all seasons stocks , tulips , etc. , produce flowers which break .
" For example a pure strain of self-coloured wallflowers may become blotched or flaked .
There is evidence in the case of wallflowers that this breaking has no relation with genetical constitution , for a strain which 4\lt ; breaks " badly one year may behave in a perfectly normal manner and 1912 .
] The Oxydases of Cytisus Adami .
produce only self- ( uniformly ) coloured flowers in the following year .
Moreover growers associate the phenomenon of breaking of wallflowers with some abnormality in the development of the " wood .
" Again , in the case of stocks it is correlated with a blotched form of leaf : the plants which bear flaked flowers possess leaves with irregular light green patches .
It may be that certain tissues of the vascular system are to be regarded as secretors of oxydases which have among other functions the formation of flower pigments and that , with the oxydase-secreting mechanism out of gear , the uniformity of colour of the flowers is destroyed .
To this subject , however , we hope to return .
In conclusion we wish to express our thanks to Mr. W. Watson , Curator of the Eoyal Gardens , Kew , and to others for supplying us with material for this investigation and have to acknowledge the assistance which we have derived from a grant by the British Association .
REFERENCES .
1888 .
Darwin , Charles , 'Animals and Plants under Domestication , ' 2nd edit .
, pp. 413\#151 ; 417 .
1909 .
Baur , " Pfropfbastarde , Periclinalchimaren und Hyperchimaren , " ' Ber .
Deutsch .
Bot. Ges .
, ' 1909 , vol. 27 .
1910 .
Buder , J. , " Studien an Laburnum Adami , " ' Ber .
Deutsch .
Bot. Ges .
, ' 1910 , vol. 28 .
1912 .
Keeble , F. , and Armstrong , E. F. , " The Distribution of Oxydases in Plants and their R6le in the Formation of Pigments , " * Roy .
Soc. Proc. , ' 1912 , B , vol. 85 , p. 214 .

