BRACT-CHARACTERS 41 
cific character. So does the presence of puberulence over the 
back, or of webby hair (A. divaricatus, A. patentissimus), or stri- 
gose hair (A. patens), or of glandular hair (the Macrophylli, etc.) 
Sguarrose bracts with abruptly reflexing tips are highly distinc- 
tive of A. mirabilis, A. Herveyi, A. spectabilis, A. anomalus. Such 
squarrose tendency seems to be stimulated in the cordifolial groups 
by trampling, when the heads are thrown into an unnatural pros- 
trate condition and the bracts separate half-way down and become 
so divergent as to be reflexed (fide plants at High Bridge Park, 
"Y 
This phenomenon seems somewhat akin to the greater acumi- 
nation and outreach of teeth and leaf-apices on the part of pros- 
trated Divaricati when growing in much shade, and perhaps in both 
cases is due to the effort for light; but then why was not the 
leaf-direction similarly altered? An entirely different explanation 
must be sought for the squarrose-directed bracts of those species, 
like A. spectabilis, which produce them normally in erect growth 
in the open sun. 
Coriaceous, succulent, all-herbaceous or part-herbaceous char- 
acters are usually quite constant for the bracts of the species which 
possess them. So also of the flatness or curvature of the back ; 
or it may be keeled along the midrib; the midrib may itself be 
obsolete (A. curvescens), or a faint dark line (A. oviformis), or a 
broad, herbaceous, green band (A. vittatus). 
The dark tips of the bracts, produced by a somewhat herbace- 
ous or even pulpy expansion of the green tissue accompanying the 
midrib, are of very characteristic form in many species, elliptical in 
A. Novi-Belgii, rhomboid in A. laevis, long diamond-form, lanceo- 
late, narrowly elliptical or of various other shapes in other species. 
The middle bracts are those presenting these tips in best develop- 
ment. Bracts with some aculeate hardening above the fleshy 
" green-tip" occur among the Multiflores. 
In all broad-leaved asters the inner bracts are as a rule suc- 
sessively higher, longer, narrower, and more tapering, also 
smoother, thinner, less colored, and less ciliate. But within these 
limits each species tends to have its own norm. That norm can- 
not be reliably established however by magnifying a single head 
taken at random ; nor by taking any number of lateral heads with- 
