Note on the Morphology of the Heart of Vertebrates. 41 
make the “ Law of the Irreversibility of Evolution,’ enunciated by Dollo! and 
generally accepted by paleontologists, so puzzling to the morphologist. Using 
the longitudinal ridges as landmarks in the way already indicated, it becomes 
clear that in Lepidosiren (as in Ceratodus—Boas) the conus has—no doubt 
owing to its increase in length within the confined pericardiac space—become 
bent twice upon itself in the way indicated in the accompanying figure 
(upper part of Fig. A).” 
It will be seen that the result of this double flexure, combined with the 
fixation of the cephalic end of the conus, is that the hinder end of the conus 
has been forced into the position that it would have had if the conus had 
undergone a process of twisting in a counterclockwise direction (as seen from 
behind) through an angle equal to three right angles. It is also evident that 
were the tube figured to shrink in length, so as to straighten out the two 
kinks, such a spiral twisting would become apparent distributed through- 
out the length of the conus—the spiral being a right-handed one, as shown in 
Fig. B. If we suppose the cavity of the front end of the tube (ze. the part 
which retains its original position in relation to the body) to be divided by a 
horizontal partition into a dorsal and a ventral cavity, then that one of the 
two cavities which is at the headward end dorsal in position, will at the 
hinder end of the tube be on the right side, and the partition will now be 
vertical, its originally left-hand edge being now dorsal, and its originally right- 
hand edge being now ventral. 
In the higher Amniota, ¢.g. in a Mammal, the portion of the primitive 
cardiac tube lying between the ventricle and the headward limit of the 
pericardiac coelome, and therefore homologous with the conus arteriosus of 
the Elasmobranch, takes in the embryo the form of a gently curved tube 
which develops in its lining the usual four longitudinal endocardiac ridges. 
Of these the morphologically right (Ridge 1 of Boas) meets and fuses with its 
vis-a-vis (Ridge 3 of Boas) so as to divide the cavity by a “pulmonary 
aortic” septum. In this case the septum eventually splits, so that the 
primitive conus becomes resolved into completely separate pulmonary and 
aortic vessels. The longitudinal ridges of the Amniota are seen at their first 
appearance to describe a spiral course, so that the conus has the appearance 
of being twisted in a right-handed spiral, and the same, of course, holds with 
the resulting vessels which twist spirally round one another. It is always 
1 1893, Bull. Soc. Belge de Geologie, etc., vii, 
2 The fact that the two longitudinal limbs of the conus have slipped into a position 
somewhat dorsal to the middle limb is probably due to the pericardiac space being limited 
in width and depth as well as in length. Obviously the three limbs pack together more 
compactly than they would if the curvature were all in one plane. This relative position 
of the three limbs of the conus is of importance in connection with what follows, 
