330 H. H. NEWMAN 



production of double monsters, such as Siamese-twin blastulae 

 or gastrulae. In most cases these double embryos consist of two 

 blastulae of unequal size which together constitute a dumb- 

 bell-shaped mass with two cleavage cavities. In one case (fig. 

 12) there was found a double larva in which gastrulation had 

 occurred equally well in both the right and the left component. 

 It is fair to conclude that these are instances of a later or less 

 complete isolation of the original blastomeres of the two-cell stage, 

 bearing the same relationship to the completely separate twins 

 that human double monsters do to human duplicate twins. 



d. The individuals in which cleavage goes on typically, though 

 at a rate slower than normal, reach the blastula stage without 

 further evidences of twinning, but one who has studied the fate 

 of a sufficiently large number of such blastulae learns to differ- 

 entiate them into several groups. The first group consists of 

 solid blastulae in which the absence of a lumen renders embolic 

 gastrulation impossible. Whether gastrulation by delamination 

 or by the inward migration of endoderm cells takes place has 

 not been determined. There is no evidence that solid blas- 

 tulae ever progress beyond that stage. The second type, which 

 is quite common, is one in which the blastula is without visible 

 polarity. All of the cells are of equal size and there is no indi- 

 cation of a distinct animal and vegetal pole (fig. 13; cf. fig. 1, a 

 normal blastula) . Such blastulae undergo multiple gastrulation. 

 The surface invaginates in numerous places and a bizarre type 

 of larva (fig. 14), all wrinkles and pockets, appears, which may 

 live for days unchanged. A third type, or really series of types, 

 occurs in which a double polarity is evident. Instead of only 

 one thickened plate of endodermal cells there may be two (fig. 

 15) or even, in rare case, three (fig. 17) such plates, and gastru- 

 lation occurs at two (fig. 16) or three (fig. 13) places. The two 

 plates may be closely conjoined on one side of the blastula; 

 they may be directly opposite to each other or they may have 

 axes at right angles to each other. The result is that double, 

 occasionally triple, larvae are formed in which there is but one 

 ovoid larval body, but two, three, or more archentera. As a 

 rule, there is no difficulty in distinguishing a primary archenteron 



