436 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1912. 



amount of yellow increases and appears in broad regularly distrib- 

 uted longitudinal bands. The other half of the offspring if grown 

 on dark soil become less yellow, always, however, in close relation 

 with the opposing influence of the color of the surroundings, and 

 likewise in a regular order — in this instance as rows of spots along 

 the sides of the body. 



If the parent generation of the fire salamander be raised on black 

 garden soil (fig. 13, P row), after some j^ears it becomes largely black, 

 while the young kept upon black soil (F\ row) have a row of small 

 spots on the middle of the back. On the other hand, in young which 

 in contrast with their parents have been raised on yellow soil, these 

 spots fuse into a band. 



If we use yellow paper instead of yellow soil and begin our experi- 

 ment, as we did before, with scantily spotted individuals, then we 

 obtain enlargement, but no increase in the number of the spots. 

 If we take black paper, then we obtain a reduction in the size of 

 the spots without reduction in intensity of coloration. The young 

 bear the few spots in the middle, while the normal young from the 

 control brood in mixed surroundings at once produce an irregular 

 pattern of markings. 



Heavy moisture produces an increase of the yellow, but only in 

 the number of spots, none in the size of the spots. Numerous but 

 still small spots may be observed in the progeny put back into less 

 moist surroundings. Comparative dryness results in loss of bril- 

 liancy, but not in loss of size in the spots. The same phenomenon 

 may be observed in the progeny which is again kept moist, especially 

 when compared with the control brood which was kept under uni- 

 form conditions. 



Striped fire salamanders occur not only as fancy products of 

 breeding, but in some places (as in north German}^ and southern 

 Italy) they occur also in nature. If such examples be kept upon 

 yellow soil, then the interruptions which may occur in their stripes 

 are filled out and the completed bands become wider and send out 

 cross bridges. If, on the other hand, animals with complete stripes be 

 kept on black soil, the stripes become narrower and break up. I em- 

 ployed striped salamanders for another experiment, using some which 

 had been caught wild and some produced by artificial rearing from 

 spotted individuals. I exchanged the ovaries, grafting these of the 

 striped salamanders upon the spotted ones and that of the spotted 

 salamander upon the striped females. I can not enter upon all the 

 combinations of these experiments, especially not upon the use of 

 suitable males, which in the attaining of safe results become very 

 much involved, and would require much repetition. But the follow- 

 ing result will hardly be altered: The ovaries taken from a spotted 

 female which were implanted upon a wide striped female uniformly 



