vi HEART OF LEPIDOSIREN 377 



ventricle. It is of morphological interest to notice that this atrio- 

 ventricular ridge extends IVmn its dorsal end not directly ventral- 

 w.inls luii towards the ;iniin;irs right side, so that, if the ridge in 

 question he takm as marking an originally longitudinal line along 

 the \vall of the cardiac tuhe, it indicates that this part of the cardiac 

 tube has undergone a process of twisting like that of a left-handed 

 screw, in other words a twisting of the kind which might be expected 

 on the hypothesis that the flexure of the atrioveutricular portion of 

 the heart was originally that suggested in the discussion on p. 372. 



A second endothelial proliferation takes place along the atrial 

 wall facing that on which the atrioventricular ridge has developed. 

 The projection formed in this way grows towards and fuses with the 

 atrioventricular ridge to form the atrial septum which divides the 

 atrium into a larger right and a smaller left auricle. Owing to the 

 left-handed position of the atrioventricular ridge at its dorsal end 

 the sinus opens into the right auricle. The pulmonary vein as it 

 develops comes to open on the left face of the septum, i.e. into the 

 left auricle. 



The ventricle becomes similarly divided into a right and a left 

 chamber 1 the foundation of the septum consisting of the atrio- 

 ventricular ridge already mentioned. In this case however Robertson 

 does not describe any endocardiac proliferation vis-a-vis to the 

 ridge but says the septum is completed by muscular trabeculae 

 growing towards and eventually fusing with the edge of the ridge. 



The conus arteriosus is characterized by the development of a 

 series of longitudinal endocardiac ridges similar in nature to those 

 of Elasmobranchs. A conspicuous difference in detail is that each 

 ridge is markedly discontinuous, the portions situated in the 

 anterior and in the posterior section of the conus developing 

 independently. We may take it that the ridges were primitively 

 in the Vertebrata longitudinal and continuous and the secondary 

 discontinuity visible in Lung-fishes and also in the higher Vertebrates 

 may be associated with two probable causes : (1) interference with 

 the development of the middle region of the conus by the flexure 

 into which it is thrown, and (2) the tendency, as seen in 

 Elasmobranchs and Ganoids, for the terminal members of the 

 longitudinal rows of valves to become enlarged relatively to the 

 rest. There can be no doubt that the longitudinal ridges as we see 

 them in the Lung-fishes and the higher Vertebrates are revertive 

 rather than persistent primitive features. In other words the 

 ancestors of these Vertebrates passed through the phase of evolution 

 in which each ridge had become converted into a row of pocket 

 valves. This seems clearly indicated by the fact that the latter 

 condition holds in modern Elasmobranchs and primitive Ganoids. 

 But if so then the tendency for the terminal valves of the row to be 

 specially developed in that earlier phase of evolution may show itself 



1 The separation does not normally become quite complete in the adult either in 



the rase of atrium or ventricle. 



