May 16, 1902.J 



SCIENCE. 



767 



obvious that more abundaxit food must be 

 responsible for the greater variability of 

 domestic races. The organs, which in the 

 feral state are continually exercised in a 

 severe struggle for existence (in seeking 

 food, pursuing prey, eluding enemies, in 

 addition to other energy-consuming activi- 

 ties), do not under domestication compete so 

 closely with one another for the less needed 

 nutriment. Hence organs like the repro- 

 ductive glands, which are not so directly 

 implicated in self-preservation, are able to 

 avail themselves of more food, and this 

 should make possible, among other things, 

 more numerous combinations of the vary- 

 ing elements. That greater abundance of 

 food is thus one of the most potent, though 

 indirectly effective, causes of variability in 

 domestic races, may be regarded as an 

 established fact, in so far as we are able to 

 be certain of anything relating to this mat- 

 ter. It is a fact that the variability of these 

 races in comparison with the conditions in 

 a state of nature, has been enormously in- 

 creased, and it would be difficult to point 

 to any other factor in domestication of 

 such decisive importance to the organism 

 as surplus of food." 



As- Adlerz suggests in his paper, wild 

 animals present certain peculiarities analo- 

 gous to those exhibited by domesticated 

 forms, viz., in the enormous numerical in- 

 crease of periodic occurrence in many if not 

 in all species. This increase must depend 

 on conditions similar to those which pro- 

 duce a high rate of variability in domestic 

 forms, i. e., abundance of food and favor- 

 able meteorologic conditions. We should 

 therefore expect to find a greater amplitude 

 of variation as well as a greater number of 

 individual variations during such increases 

 than during periods of more limited in- 

 crease. This is supported by the facts, as 

 sho'wn by Adlerz 's observations on two such 

 numerical increases in a butterfly {Polyom- 

 mahis virgmirece) : 



"These two increases in number were 

 observed in the province Medelpad in cen- 

 tral Sweden. The first occurred in 1896. 

 That year the butterflies were seen to in- 

 crease greatly during July till, by the 

 middle of the month, they outnumbered all 

 the remaining species of diurnal Lepidop- 

 tera. At the same time were observed a 

 great number of a female variety not to be 

 found in any other part of the country. 

 This variety was distinguished by a series 

 of light blue spots within the reddish-yellow 

 band across the upper surface of the hind 

 wings. The number and clearness of the 

 spots varied greatly. The highest number 

 was five, and from the varieties with the 

 full number clearly developed, to the in- 

 dividuals of the dominant form, which had 

 no spots at all, an uninterrupted series of 

 transitional variations could be established. 



"During the following year the species 

 was not conspicuously abundant. Single 

 individuals of the above described variety 

 also appeared, but they were both abso- 

 lutely and relatively much rarer than 

 among the increased number of the pre- 

 vious year. Moreover, no such marked 

 variations were seen as on that occasion. 

 Hence the statement that the amplitude of 

 variation during 1896 was greater, would 

 appear to be admissible. 



"During the past summer, 1901, the same 

 species of butterfly reappeared in great 

 numbers, and again, as in 1896, during the 

 latter half of July the number of indi- 

 viduals was seen to exceed that of all the 

 other diurnal Lepidoptera. And again the 

 above-mentioned variety appeared in great 

 numbers. By counting all the females 

 among the numerous individuals that had 

 settled on several large tansy patches, I 

 found that many more than half showed 

 the variation in question to a greater or less 

 extent. 



' ' These two numerical increases thus sup- 

 port the conclusion derived from theoretical 



