692 Wisconsin Academy of Sciences, Arts, and Letters. 
regions in which the plastid is thicker are all parts of one 
plastid and not separate plastids, as Schimper supposed. 
The fact that after a starch grain is partially dissolved 
in the plastid, leaving an irregular and corroded outline, new 
layers are added filling up the depressions in the corroded sur¬ 
face, is offered as further evidence by Schimper (36, p. 187) 
in favor of the theory of external growth. It was found that 
in the development of the storage starch grains of Dolichos 
lafolab , periods of solution were followed by periods of growth 
of the grain in the plastid. After a period of solution, the 
outline of the remnant of the starch grain is very irregular, 
but the new layers added fill up the depressions and form reg¬ 
ular layers around the corroded fragment. 
Salter describes in considerable detail the development of 
a potato starch grain. There first appears, when the grain is 
stained with the triple stain, a rounded white body in the cen¬ 
ter of the plastid. In this body there soon appears a dark 
violet dot at the center. A narrow pale violet zone is next 
seen surrounding the center point. This zone darkens in color 
at the margin and becomes the first lax lamina, eccentricity 
being already indicated by the fact that it makes its appear¬ 
ance first on the side turned toward the thicker part of the leu- 
coplast. The second lamina appears in the same manner as 
did the first, the peripheral portion remaining colorless. 
I have examined starch grains from a potato for the stages 
in growth and find that while the first appearance, that is of 
the colorless body in the plastid, is as Salter describes, there 
appears to be no eccentricity of the grain shown when the first 
violet layer forms, and indeed several violet layers usually 
form before there is any tendency toward eccentricity (PI. 
XXXVIII, Fig. 14). Later in the development, the plastid 
collects more at one side of the grain and eccentric layers are 
formed. 
In suitable Canna material stained by the triple stain of 
Flemming, we find the different stages in the development of a 
starch grain very clearly shown (Figs. 2-5). Certain of these 
show no signs of lamination, others no larger in size show one 
or two pale violet circles but no broad violet layers (Fig. 3). 
