2S6 
BULLETIN OF THE BUREAU OF FISHERIES. 
always so. Distortions arise which may have been caused by a pinch and arrest of 
growth while the claw was soft or by injury to the stump. In the latter case the 
member might be only partially restored, and unequal growth would account for 
the defect. 
A small budlike swelling is sometimes seen near the apex of either division of the 
claw, and it formerly seemed to me improbable that this could be due to a simple 
injury since such appearances are rare, while injuries to the big claws must be excessively 
common. I further assumed that, given such an outgrowth, a progressive series of 
changes might take place with successive molts, the swollen part becoming bifid and 
eventually completely divided. To continue the account upon this basis: With the 
growth of the animal, the superadded part, whether it be upon dactyl or propodus, 
seems to be shifted at each molt farther and farther back upon the claw, and meantime, 
in most cases, to undergo fission in a vertical or somewhat oblique plane. This fission 
apparently proceeds until one or both of the supernumerary dactyls are entirely 
separated. The opposing edges of these become gradually toothed, so that each is 
almost an exact copy of the original. According to the principles laid down by Bateson, 
the part which is nearer the original joint corresponds with the appendages on the 
opposite side, that which is farthest away with those on the same side of the body. 
Many cases occur, however, which do not conform to this and apparently to no other 
rule (see 749, p. 144-148). 
Since the appearance of my earlier work referred to above, the excellent researches 
of Przibram {220-223) and Emmel have added greatly to our knowledge of this 
subject. The former has shown that in all probability monstrous growths of every 
kind result from a regenerative process following upon injury. However, such growths 
are comparatively rare and follow only upon injury of a certain kind, or upon an 
injury inflicted at a certain time with respect to the molting period, or under certain 
conditions of the animal which are not fully understood. 
Przibram found that when an injured leg was retained duplication of the part 
might arise through a division of the regeneration rudiment, as in vertebrates, and it 
was further shown by Miss Reed that when a leg of the hermit crab is thrown off, if the 
base is split lengthwise so as to divide the nerve, there often appear two new legs, each 
connected with one end of the nerve. It would thus appear that duplication of a limb 
is subject to the will of the experimenter, and that duplicated parts may often arise in 
nature through an accidental injury to the nerve rudiment. Further, in 1905 Zeleny 
(290) obtained by experimental means the regeneration of a double chela in the fiddler 
crab. Two cases where duplication of parts of the big claw followed directly upon 
injury to the claw itself or to a regeneration bud have been recorded by Przibram {223)] 
the first concerned a specimen of Portunus hastatus, which suffered in an aquarium the 
loss of both points of its big right claw in an irregular manner, and regenerated within 
three months; after molting, the dactyl became doubled, while the propodus was 
unchanged. The second case arose through an artificial division of a normal regenera- 
tion bud of the last walking leg of a Carcinus moenas. The operation was performed 
with fine scissors on May 14, 1901, and after the molt, which occurred on June 2, the 
protopodite showed two separated dactylopodite buds. Since this animal died on 
