are proof of conjugal devotion is the fifth of the mistakes to be corrected in this chapter.  These stories were written by men, selfish men, who intended them as lessons to indicate to the women what was expected of them.  Were it otherwise, why should not the men, too, be represented, at least occasionally, as devoted and self-sacrificing?  Hector is tender to Andromache, and in the Sanscrit drama, Kanisika’s Wrath, the King and the Queen contend with one another as to who shall be the victim of that wrath; but these are the only instances of the kind that occur to me.  This interesting question will be further considered in the chapters on India and Greece, where corroborative stories will be quoted.  Here I wish only to emphasize again the need of caution and suspicion in interpreting the evidence relating to the human feelings.

WIVES ESTEEMED AS MOTHERS ONLY

So much for the feminine aspect of conjugal devotion.  In regard to the masculine aspect something must be added to what was said in preceding pages (307-10).  We saw there that primitive man desires wives chiefly as drudges and concubines.  It was also indicated briefly that wives are valued as mothers of daughters who can be sold to suitors.  As a rule, sons are more desired than daughters, as they increase a man’s power and authority, and because they alone can keep up the superstitious rites which are deemed necessary for the salvation of the father’s selfish old soul.  Now the non-existence or extreme rarity of conjugal attachment—­not to speak of affection—­is painfully indicated by the circumstance that wives were, among many races, valued (apart from grossly utilitarian and sensual motives) as mothers only, and that the men had a right, of which they commonly availed themselves, of repudiating a wife if she proved barren.  On the lower Congo, says Dupont (96), a wife is not respected unless she has at least three children.  Among the Somali, barren women are dieted and dosed, and if that proves unavailing they are usually chased away.  (Paulitschke, B.E.A.S., 30.) If a Greenlander’s wife did not bear him any children he generally took another one. (Cranz, I., 147.) Among the Mexican Aztecs divorce, even from a concubine, was not easy; but in case of barrenness even the principal wife could be repudiated.  (Bancroft, II., 263-65.) The ancient Greeks, Romans, and Germans, the Chinese and Japanese, could divorce a wife on account of barrenness.  For a Hindoo the laws of Manu indicate that “a barren wife may be dispensed with in the eighth year; one whose children all die, in the tenth; one who bears only daughters, in the eleventh.”  The tragic import of such bare statements is hardly realized until we come upon particular instances like those related by the Indian authoress Ramabai (15):