PRODUCTION OF TWINS AND DOUBLE-MONSTERS 347 



producing twins involves retardation, physiological isolation of 

 parts of the egg or embryo, the origin of separate growing points, 

 or apical points, and the consequent doubling of normally single 

 embryos or structures. What, in Patiria, we may ask, corres- 

 ponds with the renewal of developmental vigor supposed to be 

 the result of primary placentation in the retarded armadillo 

 embryo? In every case where good results were obtained and 

 twins developed beyond an early gastrula stage, these results 

 were due to separating the living embryos from dead eggs and 

 placing them in fresh sea-water. The vessels containing twin 

 larvae were relatively roomy for the number of larvae present, 

 and water was changed at regular intervals during the periods 

 of observation. Doubtless, then, the rejuvenation of the re- 

 tarded embryo, which permitted a continuation of development 

 in otherwise doomed eggs and embryos, resulted from suddenly 

 improved environmental conditions. Many larvae were short- 

 lived, showing an inability to recover from the lethal changes 

 that had set in, others recovered to become very abnormal 

 single embryos, but a large number of twin larvae were nearly 

 always present whenever eggs that had been retarded were 

 reinvigorated by improved conditions. It is probable, also, that 

 recovery in the case of hybrids, and possibly in other cases, was 

 not strictly due to improved external conditions, but was a 

 matter of internal adjustment or of acclimation. It is hardly 

 possible from our data to decide between these alternatives in 

 any single case. Physiologically speaking, however, acclimation 

 and recovery are so closely interrelated that it seems probable 

 that they are in most cases much the same process, or at least 

 phases of one process which for want of a better word may be 

 called rejuvenescence. What appears to happen in all cases of 

 twinning is a primary dedifferentiation of the original apical 

 point and a subsequent redifferentiation of two or more apical 

 points in the place of one. There are numerous evidences of 

 dedifferentiation, partial or complete, in the retarded embryos 

 and larvae of Patiria, and there are equally numerous instances 

 of redifferentiation of plural apical points in an embryo formerly 

 with a single apical point. The simplest expression of this process 



