106 _ 



mutilation of these embryos was very extensive, reaching from the 

 cloaca forward to the middle region of the body, as in Fig. 3. 



Fig. 3 shows the injured region, on the right side of the embryo, marked by 

 three crosses. 



Almost immediately after the operation the pigmented ectoderm 

 cells began to close over the wound always pushing up dorsally from 

 the ventral side. 



After the wound had healed, the scar always appeared in these 

 embryos as a very heavily pigmented line, lying just below the ventral 

 border of the myotomes, between the myotomes and the yolk-region. 



Many embryos died during the first two weeks after the oper- 

 ation so that at the end of two months but forty healthy embryos 

 remained. Within three to four weeks after the operation all the 

 embryos had acquired distinct rudimentary limb-buds; while within 

 seven to eight weeks afterward the more vigorous sylvatica embryos 

 had developed perfect posterior legs. In all of these larger embryos 

 the right limb was perfectly formed and exactly resembled the limb 

 on the left (uninjured) side of the body. The only external evidence 

 of injury to any of these embryos is to be seen occasionally in a 

 spinal curvature that is due to injuring the medullary tube at the 

 operation. 



Kochs ') thought from his extirpation experiments on the eyes 

 and limbs of salamanders that a regenerated organ is generally 

 smaller than the normal one. He says: — "Meines Wissens existirt 

 keine Angabe darüber, dass eine Extremität völlig normal in Form 

 und Grösse regenerirt wäre". I habe found no evidence that the 

 regenerated limb is ever smaller than the normal one, when the extirp- 

 ation is made while the limb is merely a rudimentary structure. 

 My experiments differ from Kochs', however, and also from those 

 of Barfurth 2 ) in that they were made on very young embryos in 

 which differentiation of tissues had not begun in the limb-region. The 



1) Kochs, Archiv für mikrosk. Anat., Bd. 49, 1897. 



2) Barfurth, Archiv f. Entwickelungsmechanik, Bd. 1, 1895. 



