332 PHYSIOLOGICAL DEVELOPMENT. 



so fast are the efferent currents of blood, prevented from 

 escaping laterally, obliged to move from the centre towards 

 the circumference ; and so fast also does the less- developed 

 set of channels become, of necessity, occupied by afferent 

 currents, When, by a parallel increase of definiteness, the 

 lacunae and irregular sinuses through which the afferent cur 

 rents pass, become transformed into veins, the accompanying 

 disappearance of all stagnant or slow -moving collections of 

 blood, implies a further improvement in the circulation. 



By what agency is effected this differentiation of a definite 

 vascular system from the indefinite peri- visceral sac ? No 

 sufficient reply is obvious. The genesis of the primordial 

 heart is not comprehensible as a result of direct equilibration ; 

 and we cannot readily see our way to it as a result of in 

 direct equilibration ; for it is difficult to imagine what favour 

 able variation natural selection could have seized hold of to 

 produce such a structure. A contractile tube that aided 

 the distribution of nutritive liquid, being once established, 

 survival of the fittest would suffice for its gradual extension 

 and its successive modifications. But what were the early 

 stages of the contractile tube, while it was yet not sufficiently 

 formed to help circulation, and while it must nevertheless have 

 had some advantage without which no selective process could 

 go on ? This part of the question we must leave as at present 

 insoluble. To another part of the question, how 



ever, an answer may be ventured. If we ask the origin of 

 those ramifying channels which, first appearingas simple chan 

 nels, eventually become vessels having definite walls, a reply 

 admitting of considerable justification, is, that the currents of 

 nutritive liquid forced and drawn hither and thither through 

 the tissues themselves initiate these channels. We know that 

 etreams running over and through solid and quasi-solid inor 

 ganic matter, tend to excavate definite courses. We saw 

 reason for concluding that the development of sap-channels 

 in plants conforms to this general principle. May we not 

 then suspect that the nutritive liquid contained in the tissue 



