Graft-Hybrids. 
213 
analogous with the earlier known graft-hybrids. Thus, among their 
axillary and adventitious shoots, one not uncommonly finds complete 
reversions to one of the pure species, as well as chimaeras, in 
which the pure species is united with the intermediate type ; or 
again, one intermediate type may give rise to one or more of its 
fellow types. In this last respect they of course differ from those 
graft-hybrids of which only one form exists, but an analogous case 
is to be found in the Cratcego-mespilus hybrids of Bronvaux, of 
which two distinct types exist. 1 
Winkler’s results, which have been repeated and extended to 
other species of Solauum by Heuer 2 , have established the fact that 
shoots of an intermediate type may be produced as a result of the 
vegetative process of grafting. The clue which led to an explanation 
of the phenomenon came from Baur, as a result of his experiments 
on Pelargonium zonalep 
Baur crossed a white-leaved race of this species with a green¬ 
leaved race. The hybrid seedlings were found to consist of green 
and white parts arranged in a mosaic. Some of these seedlings 
give rise to adult plants which are sectorial chimjeras of green and 
white parts, exactly analogous to Winkler’s sectorial chimgeras of 
Tomato and Nightshade. The boundary between the white and 
green parts in the stem of a sectorial chimaera is often radial, but 
in some cases it follows a line such that one or two cell-layers of, 
say, white tissue overlie the internal green tissue. Such plants 
bear leaves with white margins and, as their anatomical structure 
shows, they are chimteras in which the boundary between the two 
components is not anticlinical but periclinal. 
When Winkler described the first Solanum graft-hybrid which 
he obtained, Baur 4 suggested that it might be a percilinal chimaera. 
This, and the subsequent suggestion that other so-called graft- 
hybrids might be of the same nature, has since been proved to be 
well founded. 
Winkler 5 has shown that four of his types are periclinal 
chimteras ; the four types represent the combinations to be obtained 
when either the epidermis alone or the two outermost layers may 
be derived from either species. Macfarlane’s work suggested, what 
Baur G and Buder 7 have since confirmed, that, so far as one can 
rely upon anatomical evidence alone, Cytisus Adami consists of an 
epidermis of C. pnrpureus surrounding internal tissue of Laburnum 
vulgare. Baur has shown that in the two Cratcego-mespilus hybrids 
respectively, one or two peripheral cell-layers are derived from the 
> Noll (8). 
2 Heuer (7). Heuer obtained successful results with S. Melongena 
and with S', dulcamara, as well as with the two species used 
by Winkler. He gives a photograph (p. 435) of a sectorial 
chimaera of S. lycopersicum and S. nigrum bearing lateral shoots 
of five kinds—the two pure species and three intermediate 
types. 
3 Baur (1). 
4 Baur (2). 
3 Winkler (15). 
0 Baur (3). 
7 Buder (4.) 
