PALM AND SOLE STUDIES. 225 



need not be watched longer than three or four years to get a set 

 of satisfactory prints. These objections have, however, some 

 validity as applied to my definite claims at the time when I first 

 made them (1902, 1904, 1908), and I am only glad that my 

 hypotheses of that time have been so ably and satisfactorily 

 substantiated by the results of the investigation of the armadillo. 



That the formation of multiple embryos in man and the 

 armadillo is due in both cases to a similar cause seems to me 

 undoubted, and was similarly considered by Patterson, who writes 

 (1913, p. 560): "There has been considerable speculation as to 

 how 'indentical twins' and similar types of development have 

 arisen, and I believe that these studies on the development of 

 the armadillo will at least indicate how these phenomena may 

 have come about." He further considers it probable that 

 "specific polyembryony in the Dasypodidse began by the for- 

 mation of a set of twins, perhaps at first a sporadic case of gemel- 

 liparous development such as probably occurs in the production 

 of duplicate twins in the human species." It is needless to state 

 that I, too, think the same, and because of this belief can go 

 farther and state that whatever is learned in the future concerning 

 the actual causes of polyembryony in the armadillo will solve also 

 the same question concerning the causation of human duplicate 

 twins and double monsters. 



To close this subject with a pure speculation, suggested by the 

 work upon the armadillo, Newman and Patterson find (1915) 

 that the quadruplets may be arranged in pairs, I + 11, and III 

 + IV, and that the identity of scale configuration is much greater 

 in the two members of a single pair than in members of different 

 pairs. Is it conceivable that human duplicates may bear to 

 each other these two degrees of similarity found here, and that 

 the few cases, like my Nos. VII and XIII, where the palm and 

 sole similarity was not so great as was to have been expected 

 from the facial resemblance and correspondence in sex, may be 

 explained by a relationship such as would be found in the arma- 

 dillo by comparing II with III, or I with IV? This might 

 involve the early abortion of additional embryos, or might deal 

 with the question of the two vs. the four-celled stage, as suggested 

 by Newman and Patterson (1909, p. 186) that "the two in- 



