46 VARIATION IN ASTER 
value, we might infer for Aster an ancestry with broad non-cor- 
date, spatulate or subcircular leaves, little serrate or entire. 
Granting the above, it would then become possible to explain 
as reversions certain curious subcircular or broad-spatulate stunted 
cauline leaves which occur perhaps among all asters as occasional 
intercalations among the normal leaves. Such is the case with an 
intercalary form of A. suócymosus where, after two or three fully 
developed cordate basal leaves, the stem intercalates a little 
roundish or oval entire leaf of the primordial type and then begins 
again with the modern cordate type ; the nodes and the stem itself 
being as usual above and below. Among numerous examples 
of this species all apparently in the same conditions of shade and 
nutrition, perhaps two thirds developed this intercalary leaf. 
Such an intercalary leaf is sometimes produced by the allies 
A. Schreberi, A. glomeratus, etc. 
Very similar is the phenomenon of the interrupted-form 
assigned here to A. Jussiei, with which it grows. A little oval 
entire leaf appears suddenly halfway up the stem after a full series 
of large cordate leaves. But here the cordate series is not resumed, 
the oval leaf-type gradually enlarging and changing into greatly 
augmented axiles and producing a leafy top. 
A whole series of arrest-forms follows next for comparison, 
forms occasional in most species and becoming typical in A. 
biformis, in which sudden arrest of the series of proper cauline 
leaves is followed by a continuously-small subcircular or broad- 
spatulate series of little leaves without much increase in size up- 
wards. Does the leaf-form assumed under this arrest of develop- 
ment betray a trace of the ancestral type? Does the arrested leaf 
stop at the stage of development from whence the modern leaf- 
type has sprung ? 
But another interpretation of all these facts may be offered. 
The primordial leaf may owe all its characters to suppression of 
growth due to rapid passing of the growth-impulse into the bet- 
ter-illuminated higher nodes. The primordial leaf may stand in 
the succession of rapid and unfinished partial recapitulations of 
ancestral stages, but may yet fail to assume any definite or mature 
ancestral form ; it is so brief a growth. Like the first little scale- 
like rudimentary leaves on the seedling aster, so the larger first 
