248 PERCY SLADEN TRUST EXPEDITION. 



Tio stomodseum, in the latter, as the figures of Lankester and "Willey (1890) and Legros 

 (1898) clearly demonstrate. But the great and iiupoi-tant change takes place when a 

 l)elagic existence is abandoned for life in the sand. This necessitates the formation of 

 an atrium, primarily, on Goldschmidt's, as on Boveri's view, in order to make possible 

 an increase of respiratory surface in the presence of a hody musculature extending far 

 down towai'ds the ventral edge. An atrium having l)een formed and beginning to grow 

 in towards the interior, a new series of gill-slits, in no way homologous with the old, 

 arises on tlie right-hand side. Both series increase in lateral extent, and, encroaching 

 on the mouth, force it to abandon its lateral position for a new and topographically 

 symmetrical one anterior to the gill-slits. These are then enabled to follow suit, and 

 take up a symmetrical position, equally coenogenetic with that of the mouth, on either 

 side of the originally asymmetrical branchial artery. With the continued dorsal growth 

 of the gill-slits the "pars nutritoria " becomes reduced to an epibranchial groove, and 

 the endostyle, losing its original function, follows the prevailing fashion of symmetry 

 and glows back between the two rows of gill-slits *. Finally, the velum and buccal 

 tentacles appear as adaptations to the new conditions of nutrition, and the evolution of 

 Aviphioxiis as we know it is thus completed. 



This reconstruction ol' phylogeny, agreeing almost point for point with the course of 

 ontogeny in BranchiGstoma lanceolatum, Goldschmidt considers to offer a more satis- 

 factory explanation of its development than any other which has been proposed. But it 

 can hardly fail to strike one that this agreement is almost too perfect to be trusted. 

 We have learned, since the earlier days of the "Biogenetic Law," to expect in the 

 individual at best a highly modified and obscured representation of the racial history. 



Moreover, one step in the course of evolution here traced seems to me so exceedingly 

 improbable as seriously to vitiate the whole tiieory : this step is the establishment of a 

 secondary symnn^try between the two rows of non-liomologous gill-slits in place of the 

 supposed jjrimary symmetry of the originally existing series. To restate the supposed 

 course of events: — Aniphioxides possesses a median unpaired series of slits, displaying a 

 bilateral symmetry about the sagittal plane : to these are subsequently added a second 

 series, in no way homologous with them, essentially asymmetrical organs, arising on one 

 side only : the first series now acquires jx structure identical with that of its newly-formed 

 partner, and a rotation of the pharynx brings the two into symmetrical positions on 

 either side of a new sagittal plane. Surely this is a history unparalleled elsewhere in 

 the animal kingdom ! Cases are numerous enough in which an organ of one side grows 

 over on to the other and establishes apparently symmetrical relations about the sagittal 

 ])lanc — the development of the hydrocoele in Ecbinoderms may be cited as an instance: 

 tliere are none that we can call to mind of a primitively median and a primitively 

 asymmetrical organ, mucb less whole series of organs, secondarily establishing sym- 



* Goldschmidt suggests that the endostyle acquires the function of conducting food-particlcs backwards between 

 the gill-slits. It would seem more probable that (as in Ascidians) its function is to secrete a cord of mucus which 

 shall involve the food-particles, and lead it forwards to the peripharyngeal bands. Andrews has observed the food 

 to pass backw ards along the latter in Asyminelroii luaiyanum. 



