i ORIGIN OF THE MESODERM 55 



making use of any developmental phenomena except those observed 

 in typical trunk segments as a basis for speculations upon the evolu- 

 tionary origin of the mesoderm. 



It has, further, to be borne in mind that observations upon the 

 development of the mesoderm in its early stages have to be made by 

 the method of serial sections, and that in the interpretation of such 

 sections the liability to error becomes greatly increased if the sections 

 are not exactly in one of the three following sets of planes — (1) 

 transverse to the morphological axis, (2) " horizontal," and (3) parallel 

 to the sagittal plane. This type of technical difficulty is in many 

 Vertebrate embryos most marked in the head and tail regions. 



For these reasons it seems safest, in considering generally the 

 ontogenetic development and the probable evolutionary history of 

 the mesoderm, to ignore all observations except those made on 

 typical trunk segments between the level of the otocyst in front and 

 of the anus behind. This will accordingly be done in what follows. 



It is agreed by the majority of students of Vertebrate embryology 

 that the most nearly primitive condition of the mesoderm known to 

 occur in the embryos of Vertebrates is that seen in Amphioxus, where 

 it consists for a time of a series of endodermal pockets, converted 

 later into closed sacs, upon each side of the body (Fig. 34, B). 



It appears fully justifiable to conclude that both of the stages 

 mentioned represent ancestral conditions in the evolution of the 

 Vertebrate mesoderm. The excretory organs of the Vertebrate, in 

 the form of paired segmentally arranged tubes, afford in themselves 

 strong evidence that at one time the Vertebrate coelome was in the 

 form of isolated segmentally arranged chambers. 



In the case of Amphioxus the segmented character of the meso- 

 derm persists only dorsally, the ventral portions of the successive 

 segments becoming fused together so as to give rise to a continuous 

 unsegmented splanchnocoele or peritoneal cavity. 



In the Craniata the smallest departure from the condition in 

 Amphioxus is seen in such comparatively primitive forms as Lam- 

 preys, Crossopterygians and Lung-fishes. In these a solid continuous 

 mesoderm rudiment becomes split off from the endoderm on each side, 

 remaining for some time continuous laterally with the endoderm 

 (Fig. 40, B, C, p. 65). In the outer or lateral part of this mesoderm 

 rudiment the segmentation, which even in Amphioxus was only 

 temporary, never makes its appearance. The dorsal portion does 

 segment but the segment is a solid block of cells in which a cavity 

 only appears later on. It is fairly clear that these mesoderm seg- 

 ments, except for the fact that they are continuous in their ventral 

 portions and that they are at first solid (a modification of develop- 

 ment which is very common in hollow organs), agree closely with the 

 segments of Amphioxus and that they are homologous structures 

 merely somewhat modified from the primitive condition met with in 

 Amphioxus. 



In endeavouring to institute a more precise comparison of the 



