THE CAMBIUM AND ITS DERIVATIVE TISSUES 427 
ings of the connecting spindle fibers have been questioned by various 
investigators. Treub (1878) held that the cell plate arises from free 
cytoplasmic granules which migrate into the equatorial region of the spindle, 
and Zacharias (1888) contended that the spindle is of nuclear, and the cell 
plate of cytoplasmic, rather than of kinoplasmic, origin. Mottier (1900) 
found that in certain of the higher plants the connecting fibers do not 
thicken appreciably in the equatorial region, and do not lie sufficiently 
close together to enable the slightly thickened middle parts to meet and 
fuse. He concluded from this "that the cell plate is formed by a homo- 
geneous plasma which is conveyed to the cell plate region and deposited 
there by the connecting fibers." Strasbiirger's generalization concerning 
the kinoplasmic origin of the cell plate was ably defended by Timberlake 
(1900), who concluded, however, that the nucleus "is the center of metabolic 
processes concerned in the production of the kinoplasm." 
Strasburger (1880) and Went (1887) noted that the connecting spindle 
fibers appear to increase in number prior to the formation of the cell plate, 
and the former investigator subsequently suggested that this increase might 
be due to a splitting of the original fibrillae. Timberlake endeavored to 
prove that the lateral expansion of the spindle is not due to the formation 
of new fibers, but to the separation and enlargement of the original connect- 
ing fibers, coupled in certain cases with the addition of peripheral threads, 
formed from fibers which radiate from the daughter nuclei. 
Hof (1898) and Nemec (1899) showed that after the cell plate is com- 
plete the connecting fibers appear to be "drawn in" toward the cell plate 
and their place taken by granular cytoplasm. Timberlake stated that 
"concurrent with the growth of the cell plate in extent and the disappearance 
of the fibers from a central portion of the spindle the nuclei come to lie 
nearer the cell plate." He concluded that "all of the fibers which form 
cell plate elements are completely used up in the growth of the cell plate," 
and that "during the period of growth the cell plate may so shift its position 
as to lie in a plane different from that in which it was first formed." 
The formation of the cell plate in cambial initials is significant in these 
connections. There is an enormous increase in the number of " peripheral " 
fibers during cytokinesis. As in the case of the radial fibrillar systems, which 
Harper and Dodge (1914) have shown to function in the formation of the 
capillitium of certain Myxomycetes, the kinoplasmic fibers have no visible 
connection with the nuclei. They are so far removed from the nuclei 
(figs. 20, 21) during the later stages of cytokinesis that one is inclined to 
question Timberlake's generalization concerning the nuclear origin of 
kinoplasm. The phenomena appear rather to justify Strasburger's con- 
tention that the accessory fibers are of cytoplasmic origin. As the original 
connecting fibers "draw in" or disappear, the nuclei tend to move towards 
the cell plate (figs. 19, 32, 45, 55), but subsequently may move away from 
it (figs. 21, 27). That the connecting fibers may thicken considerably in 
