746 The American Naturalist. [August, 
insects probably had eggs with considerable yolk and dense envelopes 
of chitin. The rapidly growing embryo, feeding upon the yolk must 
give off waste products that might accumulate in the adjacent yolk. 
Hence, it would be of advantage if the embryo could move away to 
yolk not contaminated. 
From a comparative view of similar movements in Orthoptera the 
author infers that the more complete migration movements are primi- 
tive and that the more restricted movements have been more recently 
assumed; while in other groups various stages of disappearance of 
these blasto-kinetic phenomena have been described. 
A classification of the insects based upon these movements and the 
character of the embryonic membranes would not, the author is con- 
vinced, conflict with that of Brauer, though the embryological evi- 
dence would tend to subdivide the Orthoptera or raise the value of the 
families of this group. 
These embryonic membranes in Xiphidium are very numerous and 
complicate the above movements of the embryo, are, perhaps, 
the cause of the return movement. The serosa covering the entire 
yolk becomes cut off from the folds, amnion, covering the ventral face 
of the embryo so that the embryo first sinks into the yolk free from the 
serosa, which surrounds everything, but carrying with it the amnion. 
When the return trip begins, however, the amnion is again grown fast 
to a super-jacent membrane, the indusium, and is ruptured before the 
embryo can advance. This it does by becoming evaginated from the 
pouch of the amnion, out of which, also, an amniotic liquid escapes. 
The reflexed membranes do not form the dorsal surface of the future 
insect but are thrown off; in fact this is the case, the author thinks, in 
all insects: the envelopes may be variously absorbed or cast off in dif- 
ferent cases, but in all they are to be regarded as specialized organs 
that have performed their part and are not used again in the forma- 
tion of the body wall of the insect. 
The indusium to which the amnion adheres is a most remarkable 
organ that first appears as a faint disk of cells anterior to the embryo 
when the blastopore is open. 
Later this plate of cells is for awhile connected with the head lobes 
of the embryo, which then assumes a clover leaf outline, disappearing 
when the plate becomes again free. 
This plate of cells is grown over by the serosa so that an amnion is 
Fiii over it just as over the embryo, the edges of the plate rising 
up to meet as a thin amnion beneath the serosa. This amnion is cut 
off from the serosa; it may be called the outer indusium to distinguish 
