PRODUCTION OF TWINS AND DOUBLE-MONSTERS 345 



5. The one factor in common among the three methods of 

 producing twins in Patiria has to do with a more or less pro- 

 nounced retardation of the developmental rate, accompanied by 

 a decrease in metabolic intensity. In the case of partheno- 

 genetic eggs, cleavage begins at least three hours later than in 

 normally fertilized eggs. The retardation here is extremely early 

 and the results are seen in the physiological isolation of the 

 blastomeres of the two-and four-cell stages, and the formation of 

 completely separate twin blastulae. In the case of hybrid larvae, 

 retardation is neither so early nor so severe in its onset as in 

 parthenogenetic forms, nor so prolonged in its action. The 

 consequence is that the results are not clearly noticeable until 

 the onset of gastrulation. Excluding from consideration the 

 parthenogenetic eggs that occur in normally fertilized cultures 

 the retardation of development in larvae resulting from normally 

 fertilized eggs is not marked during the cleavage period, but 

 becomes evident when the larvae hatch and swim to the surface. 

 There are no twins among those that hatch first and are earliest 

 drawn off from the surface. Among the larvae that hatch later, 

 however, there are twin larvae of several types, some simulating 

 those in hybrid cultures, others, however, such as the 'double- 

 headed' and 'double-pored' types, have no parallels in hybrid 

 cultures. These larvae show effects of retardation at later stages 

 than do twins produced under either of the other conditions 

 previously dealt with. 



One conclusion, therefore, seems to be evident, that, in Patiria at 

 least, twinning of all sorts is intimately associated with retarded 

 development. 



DISCUSSION AS TO THE CAUSES OF TWINNING 



The writer has for some years entertained a physiological 

 theory as to the causes of twinning, derived from his studies of 

 the striking case of twinning (specific polyembryony) in the nine- 

 banded armadillo. This theory was expressed most recently 

 (Newman, '17) in the volume. The Biology of Twins, though it 

 had been stated some years earlier (Newman, '13). The situa- 

 tion in the armadillo is as follows. The egg is at first single, as 



