28 DR. J. F- GEMMiLL ON [May 12, 



in this case the two bladders lie in the same plane and have 

 corresponding right and left sides. But the right side of BL' 

 is in connection with a left Wolffian duct (^Vcld\ while the 

 left side is in connection with a right Wolffian duct. Such a 

 transposition is exceedingly rare in double monstrosities. Figure 26 

 was taken from a case of symmetrical ventral union. Two 

 vents and two urinary pores were present, and as they opened 

 laterally in pairs towards opposite sides, they lay in a plane at 

 right angles to the sagittal plane of the twin bodies. This 

 arrangement has many parallels in teratology, e. cj. in cases 

 of ischiopagous double monstrosity*. It preserves the natural 

 correspondence between rights and lefts in the ducts and bladders, 

 which, as has just been seen, is inverted in the case from which 

 figure 25 of Plate III. was taken. 



General. 



With the rarest exceptions, all double monstrosities in fishes 

 are examples either of anterior duplicity or of union by the yolk- 

 sac. The explanation of this remarkable fact seems to me to be 

 as follows : — In all these cases, two centres of gastrulation form on 

 the edge of the blastoderm at a greater or less distance from one 

 another. The spreading of the blastoderm over the yolk-mass 

 goes on freely all round except at and near the primitive streak. 

 There, changes take place which lead to inci-ease in leng-th of the 

 embryonic axis, and which are interpreted by many as con- 

 crescence. If the two centres of gastrulation happen to be near 

 one another, the whole of the blastoderm edge separating them 

 eventually will be used up in the process of concrescence ; the 

 later formed parts of the embryonic axes will be drawn closer and 

 closer to one another, until in their turn the axes themselves 

 coalesce. The degree of union will be in inverse proportion to 

 the original distance from one another of the two centres of 

 gastrulation. Should the two centres of gastrulation be so far 

 apart that the middle portion of the intervening blastoderm edge 

 is not involved in concrescence but is left free to extend over the 

 yoLk-mass, the two embryonic axes will be independent along 

 their whole length, and the only structiu^es which connect them 

 will be the blastoderm and the yolk-sac. According to this view, 

 double monsters showing anterior duplicity are the result of what 

 may be called primary fusion, that is, concrescence of their 

 growing embryonic axes. 



In birds, typical concrescence can occur only during the earliest 

 stage of formation of the primitive streak, i. e. so long as the 

 groove of the sickle and knob is open. Any subsequent concres- 

 cence can take place only by the incidental drawing in and 

 utilisation of lateral blastema at the growing zone. This process 



* J. F. Gemmill in ' Journal of Anatomy and Physiology,' vol. xxxvi. p. 263. 



