292 A. F. W. SCHIMPER. 
the case, the chlorophyll-corpuscle is discoid, the localisation 
is greater,so that the formation of starch-grains, is confined to 
the equatorial zone. Chlorophyll-corpuscles of this kind 
often produce six or more starch-grains, which form a girdle 
in this plane, whereas the central portion and the flat sur- 
faces of the corpuscles are free from them (figs. 4 and 5) ; 
occasionally a minute grain may be formed in the flat 
surface. 
These peculiarities in the mode of development of the 
starch-grains from the chlorophyll-corpuscles, and in the 
mode of their nutrition, are intimately connected with 
certain peculiarities in their structure. 
The starch-grains which originate in the interior of a 
chlorophyll-corpuscle and remain enclosed within it, have 
a concentric structure (for instance those formed in the 
cortical and medullary parenchyma of certain Cacti, such 
as Cereus speciosissimus) ; in most cases such starch-grains 
remain very small and exhibit no differentiation. The 
starch grains of Vanilla planifolia' (figs. 1—3), which 
are of this kind, deserve special mention. When mature 
they are spherical completely colourless compound grains, 
consisting of hundreds of small similar polyhedral grains ; 
they closely resemble those of the endosperm of the 
Caryophyllez and of the tuber of Mirabilis Jalapa. An in- 
vestigation of their development shows that the small grains 
arise as minute points in the chlorophyll-corpuscles; they 
then increase in size, and become polyhedral in consequence 
of mutual pressure; the substance of the chlorophyll-cor- 
puscle becomes gelatinous, decreases in size, and finally 
disappears. 
Those starch-grains which originate in the second way, 
that is, in the peripheral portions of the chlorophyll-corpus- 
cles, usually attain a much greater size. Those which occur 
in the stems of Begonia, Peperomia (e.g. P. stenocarpa), 
Pelargonium, Ozalis Ortgiesu, Dieffenbachia Seguina, Costus 
Malortieanus, and less markedly those which are present in 
the stem of the Potato, are among the largest and the most 
perfectly developed ; they exhibit a very distinct differentia- 
tion of hilum and layers. Such starch-grains are all eccen- 
tric, and the side which has grown the most is, without 
exception, the one to which the chlorophyll-corpuscle is 
attached (figs. 8 a, 10 and 12). From this fact it clearly 
follows that the unequal growth of the two sides of the 
starch-grain is a consequence of an unequally distributed 
a For this material I am indebted to the kindness of Prof, E. Morren of 
iége, 
