532 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1914. 
bud-shaped outgrowth. After the second molt regenerations are pro- 
duced consisting of a short stem of undifferentiated tarsal joints and 
a slightly developed but unmistakable claw joint with its character- 
istic parts. At this time, the regenerations have not yet reached the 
differentiation shown in figure 3. At the following molts, the tarsus 
is further developed; the joints become differentiated and each de- 
velops its pair of plantula. The difference in the size of the joints 
becomes apparent, with the first tarsal joint considerably longer (com- 
pare fig. 7) than the following, quite as in the normal tarsus. A large 
tibia-like joint was further developed in four specimens between the 
point of amputation on the second antennal joint and the tarsal joints. 
My material of these more or less developed homeotic regenera- 
tions includes 20 cases. In all these the nature of the regeneration 
is determined by the presence of a claw-bearing joint or, where this 
is absent, by the presence of joints supplied with plantula. I have 
not attempted to diagnose as either antennal or tarsal joints a large 
number of cases where the regeneration consisted of a short stem 
of undifferentiated joits, and these cases are consequently not in- 
cluded in the tabulation of the material. 
In order to get a comprehensive survey of the material, it is neces- 
sary first to note its several imperfections. Specimens of different 
ages and on which different amputations had been made (within the 
first and second antennal joints) have been reared together after the 
amputations. The method of amputation was, as mentioned, imper- 
fect and it is therefore in many cases impossible to determine with 
certainty the exact place of amputation, and thus draw conclusions 
with respect to the influence of the place of amputation on the nature 
of the regeneration. At present it is impossible to determine the age 
of each specimen at the time of amputation, a factor which is of the 
greatest importance in the development of the regenerations, as these 
are dependent upon the molts like the rest of the organism. 
However, the place of amputation can be determined approxi- 
mately in most cases. As all of the amputations in these experiments 
were made on or between the first and second antennal joints, I 
have relied in these determinations upon the relative size of the 
regenerated part and the joint from which it has grown. Due to 
its larger diameter, the second joint covers the entire apex of the 
basal jot, while the more slender regenerations are generally at- 
tached with a much smaller base to the broad basal joint. A com- 
parison with a remaining normal antenna is naturally a considerable 
help in these determinations (pl. 1; figs. 1, 2). 
In the arrangement of the material, I have given special attention 
to the insertion place, structure, and size of the regenerated appendages. 
The material may be divided into two groups, according to the 
place of amputation: 
