358 Mot tier. — On Certain Plastids , with Special Reference to 
In older cells the protein granules become .very large, and the 
impression is that these result from the fusion of smaller ones (Figs. 17 a 
and 1 y b). In old cells, as in parts of the floral envelope and scale leaves, 
these larger bodies seem to be hollow, possessing a dense dark periphery 
and a colourless centre (Fig. 18 ). They are as a rule not globular, but 
elongated and irregular in outline. These bodies give a protein reaction 
with the usual microchemical tests for this substance in plant cells. 
Conopholis is a complete parasite, utterly devoid of chlorophyll, living 
upon the roots of forest trees in this locality, and the writer was much 
surprised, when studying the preparation from which Fig. 16 was taken, to 
find starch grains. At first he was in doubt as to the accuracy of the 
observation, but an iodine test of freehand sections of both fresh and 
alcoholic stems left no ground to question the presence of starch. The 
leucoplasts in this parasite are able, therefore, to synthesize starch from 
elements obtained from its host. 
Discussion. 
In an earlier publication (Mottier, 1918) the writer traced the origin 
of leucoplasts and chloroplasts from small granular and rod-shaped pri- 
mordia, which had been generally referred to as mitochrondria, chondrioconts, 
or chondriosomes. He maintained furthermore that these primordia were 
permanent organs of the cell, and emphasized the view, probably expressed 
first by A. F. W. Schimper (1880), that leucoplasts and chloroplasts were 
morphologically alike. 
It was further pointed out that, in the same cells along with the 
primordia of leucoplasts and chloroplasts, other bodies were present, 
similar in form, size, and appearance to these primordia, which did not 
develop into leucoplasts and chloroplasts. For convenience and clearness 
the term chondriosome was applied to these rods and granules that did 
not develop farther. The same term was applied to similar rods and 
granules found in cells of certain Liverworts, such as Anthoceros and others. 
The writer did not attribute to these bodies any specific function. In the 
Summary (1. c., p. 112 ) he suggested that the bodies which were designated 
as chondriosomes, for convenience, might be the primordia of plastids. 
Guilliermond (1819) seems to take exception to the above use of the 
term chondriosome, claiming that the case of the Liverwort cited was 
a special case. He emphasized the view that granular and filamentous 
mitochondria and chondrioconts are different forms of the same structures, 
and that they are of equal significance (1. c., pp. 242 - 3 ). As a matter of 
fact my use of the term chondriosome in the Liverwort applied to both 
granular and slender rod-shaped bodies that did not develop into chloro- 
plasts, just as in the case of other plants. In later publications (1920) 
Guilliermond seems to have changed his view somewhat, as he admits that 
