1890.] Probable Causes of Polygamy Among Birds. 1031 
meet the excessive demands of consequent polygamous practices, 
3d, That such prepotency, being born, so to speak, of sexual 
selection, may be restrained by unfavorable climatic conditions or 
inherent specific peculiarities in its tendency to produce such 
practices, while among species that are free from these limitations 
polygamy finds willing victims. 4th, That domestication may 
conduce to polygamy in two ways,—first, by removing these 
limitations, as in the mallard; or second, by artificial development 
of special characters not found in a state of nature, as in the can- 
ary. 5th, That the domestication of a monogamous species of 
slight sexual differences, unless first directed to the higher devel- 
opment of secondary characters, will have no influence on the 
connubial relations, as instanced by the guinea fowl. 
It may be asked, Why, then, are not many highly-colored 
species merging toward polygamy? I reply that probably many 
are. The Trochilidz, according to Salvin, are in some cases polyg- 
amous, and so are the Paradiseidz, if we may believe Lesson, 
though Wallace inclines to doubt the fact. More intimate 
acquaintance with these families during the breeding period will 
possibly reveal indubitable proof that they contain many exam- 
ples of a habit which, as is shown, is developed only in connec- 
tion with extraordinary sexual secondary characters. 
As was mentioned in the third of the above conclusions, the 
tendency towards polygamous unions may be checked by 
many circumstances of a physical or mental nature, or 
it may be nullified by conditions of environment. Among 
many highly-plumaged groups we find less quarrelsome disposi- 
tions than is common with the majority. This may be caused 
by climatic or constitutional influences, which, however, did not 
prevent development of secondary characters by sexual selection, 
yet restrained in great degree the spirit of rivalry and consequent 
destruction of high-tempered males commonly attendant on such 
development. Again it appears that the quarrelsome disposition 
is powerless in other families to destroy much life, because of the 
ability of males to escape each other when defeated, which, coupled 
with the uniform distribution and individual independence of the 
sexes in species of arboreal habits, enables conqueror and con- 
Am. Nat.—November.—4. 
