INHERITANCE OF TEMPERAMENT. 121 



influence that makes for a greater or less display of emotional qualities 

 in the individual. Just what we shall have an impulse to do is deter- 

 mined by numerous factors; but the general nature of our reactions, 

 whether violent or repressed this is determined by the hereditary 

 nature of our temperaments. 



XII. SUMMARY. 



Problem. What hereditary factors determine constitutional tem- 

 perament and its expression in mood? What factors determine that 

 one person shall be prevailingly elated, another depressed, and another 

 still show, at different times, both of these conditions? 



Definitions. The condition of activity above the average is called 

 "hyperkinesis;" that below the average is called "hypokinesis;" and 

 each occurs in two grades. From the old psychology is borrowed the 

 terminology of temperament choleric, nervous, normal (calm and 

 cheerful), phlegmatic, melancholic; and any one of these may be com- 

 bined with any other. 



Hypothesis. The following hypothesis is proposed: There is in the 

 germ-plasm a factor, E, which induces the more or less periodic occur- 

 rence of an excited condition (or an exceptionally strong reactibility 

 to exciting presentations); its absence, e, results in an absence of 

 extreme excitability (calmness). There are also a factor C, which 

 makes for normal cheerfulness of mood, and its absence, c, which 

 permits a more or less periodic depression. Moreover, these factors 

 behave as though in different chromosomes, so that they are inherited 

 independently of each other and may occur in any combination. 



The method is the study of the progeny of 146 matings of two persons 

 who are fairly well described themselves and of whom the tempera- 

 ments of the parents usually, and of certain of their children always, are 

 known. To these parents are ascribed the most probable zygotic 

 formulae based on their behavior and the most probable gametic for- 

 mulae of their parents; then from the parental formulae is calculated 

 (table B) the probable frequency of the temperaments in the offspring, 

 and a comparison is made between the calculated and the observed 

 frequencies (tables C and D). 



Results. The hypothesis is supported by the following evidence : 



(1) By the close relation between the actual and the calculated 

 frequency of the various classes among the offspring of parents of known 

 zygotic constitution (tables C and D). The two series follow each 

 other fairly closely, except for an excess of actual normals and near- 

 normals, and an excess of manic-depressive offspring probably due to 

 the fact that our material was selected to secure as large a proportion 

 as possible of them (p. 89). 



(2) By the fact that only 0.95 per cent of the offspring do not con- 

 form to hypothesis; that the hypothesis works! Actually, of 133 off- 



