BULLETIN 824, U. S. DEPARTMENT OF AGRICULTURE. 
many of which were elongated, having oblique or pointed ends and 
being joined in the manner of prosenchyma cells. These were usually 
found as fragments in the powder, and could be recognized by their 
walls pierced with narrow canals. He also observed these fragments 
to be much more numerous in the Persian than in the Dalmatian 
powder. He explained this by the fact that the greater portion of 
the very rigid, greenish, involucral scales (with the exception of the 
dark, reddish-brown scarious margin) consisted of sclerenchyma cells. 
Numerous hairs of a very characteristic structure were found on the 
outer surface and along the membranous edges of the scales of the 
Dalmatian flowers and on the flower stems as well. Each of these 
hairs consisted of a long cell with attenuated ends, placed horizontally 
on a 1 to 3 celled stalk arising from the epidermis. The terminal hori- 
zontal cell was bent and twisted in various ways, rather hooked at the 
end and forming feltlike layers, especially on the outermost scales. 
Schrenk detected few hairs in the Persian flower insect powder 
which he examined. The flowers of Chrysanthemum roseum which 
he subsequently examined were almost entirely glabrous, with the 
exception of the hairs found where the stem widens into the recep- 
tacle, as well as at the base of the outermost scales. The hairs were 
of the same structure as those found on Chrysanthemum cinerarise- 
folium, only the terminal cells were much longer. He considered the 
papillae covering the upper epidermis the most conspicuous among the 
fragments of the marginal corolla. These were not regarded as 
diagnostic, since the petals of other related species are similarly con- 
structed. Stomata were remarkably numerous on the lower side of 
the marginal corolla. He did not find the insect flowers raised in 
California (Buhach), which belong to Chrysanthemum cinerarise- 
folium, any different in structure from flowers grown in their native 
country. 
Kirkby (156) and Verneau (285) have called attention to one dis- 
tinction which they believed might aid in identifying the Persian 
powder. They found the papillae of the ray florets to be larger than 
those of the Dalmatian florets, thickened somewhat more at the apex, 
and with sides making a wider angle. Malfatti (186) has gone into 
the description of the Caucasian {Chrysanthemum roseum) flowers 
quite extensively, figuring the various parts of the flower head. 
Siedler (258) describes the physical characteristics of powders made 
from different parts of the flower. 
Microscopic Examination op Insect Powder. 
Before attempting the critical examination of a commercial insect 
powder, the microscopist should become thoroughly familiar with 
the various parts of the insect flower head, in the whole as well as 
in the powdered condition. It has been found advantageous to sep- 
