FACTS OF INHERITANCE 165 



atra) produces at birth two fully formed terrestrial 

 young 38-40 mm. in length. 



(c) Kammerer kept S. maculosa in the cold, 

 and got it, after a few pregnancies, to produce only 

 two young ones, as in S. atra. 



{d) He kept S. atra in a warm place with plenty 

 of water, and got it to produce 3-9 aquatic larv83, 

 thus approaching the condition in S. maculosa. 



(e) The offspring of the Salamanders thus 

 treated (c and d), became sexually mature when 

 three and a half years old in conditions normal 

 to S. maculosa. The offspring of (c) gave birth 

 (1) to very advanced larvae, 45 mm. long with 

 much-reduced gills, metamorphosing several days 

 after, or moderately advanced aquatic larvae 

 40 mm. long, with large gills ; or (2) to small 

 larvae, 20 mm. long, with rudimentary gills, laid 

 on land, and metamorphosing after four weeks 

 into salamanders 29 mm. long. Thus there was 

 a partial persistence of a modified mode of re- 

 production in the absence of the modifying 

 conditions. 



(/) The offspring of (d) bore in the water 3-5 

 larvae, 33-40 mm. or 21-23 mm. in length, light 

 in colour, and possessing giUs. Thus there was 

 an augmentation of the parental modification 

 (d) in conditions which resembled those of the 

 original experiment. 



The difficulties in regard to this very interesting 

 set of experiments are : (1) they do not deal 

 with a structural modification ; (2) it is possible 

 that the experimental conditions acted directly 

 on the germ-cells in (c) and (d) ; (3) there was 

 some measure of artificiality in the conditions 

 under which, the second generation developed, 



