434 ANNUAL EEPOKT SMITfiSONIAN INSTITUTION, 1912. 



figs. 4:Cj 4:D), show an increase in dark pigment, a reduction of the 

 yolk to its complete retrogression as well as changes in the gills, 

 which become shortened, simplified, and coarser, and while usually 

 only the first of the gill arches of the skeleton bears a gill, these appear 

 on three of the free arches in the fourth generation (4D). The males, 

 probably on account of the difficulty of clasping the female in the 

 water, have developed as an adaptation coarse swellings on their 

 thumbs, and also strengthened the musculation of the arms, which 

 lend the forelimb a more inwardly flexed appearance. These are 

 external sexual characters which hold good in all toads and frogs that 

 mate in the water, but which are not normal, for the midwife toad 

 normally mates on land. 



The brooding instinct, or its absence, are peculiarities which fall 

 to the lot of the male in the midwife toad. The character and devel- 

 opment of the egg, on the other hand, are everywhere derived from 

 the females. To cross normal midwife toads with such, in which 

 the reproductive processes had been changed, presented some fas- 

 cination. In our illustration (fig. 10) we have sIiomti a normal 

 male (7 c? ) with the affixed egg chain upon its thigh, in order to indi- 

 cate that it will, if necessary, actually carry out the function of 

 brooding. The changed male (8 <? ) is larger and has left the egg string 

 with the smaller darker eggs, which are surrounded by a swelled 

 capsule, lying unnoticed alongside of him. The normal females 

 appear in the empt}'^ white field; it deposits its eggs, as we know, upon 

 the earth. The larger changed female is in the shaded field because 

 it deposits its eggs in the water. I cross in the one case (fig. 10, detail 

 fig. 7P) a normal male with a changed female. The offspring (fig. 1) 

 prove at their first reproduction to be perfectly normal; the males 

 brooding, the females depositing their eggs on land. I thought 

 nothing less than that the habit change on account of the introduc- 

 tion of the normal male in the parent generation had passed out. 

 Alas, they reappeared in the second generation (F^) almost exactly 

 in a quarter of the offspring, the remaining three-fourths of the 

 second generation being normal. An opposite crossing, a normal 

 female with a changed male (fig. 10, detail fig. 8P), yielded the 

 following results: The first generation (F^) takes exactly after the 

 male parent; that is, all the individuals exhibit the reproduction 

 changes resulting from the experiment, the females deposit the eggs 

 in the water and the males do not brood. The second generation is 

 one-fourth normal and the remaining three-fourths are changed. 



Let us now note the following scheme (fig. 11): It corresponds with 

 the experiences which have frequently been obtained in crossing 

 races of plants or animals. Namely, if one crosses a "dark" with a 

 "light" race (fig. IIP), then one of the two peculiarities alone is 

 dominant in the daughter or the first filial generation (F^), for exam- 



