554 W. D. Lang — Bernard's Biological Theories of Fossils. 



Radiolaria. As colonies formed of homogeneous units it has built 

 up certain forms such as the Sponge, the Ccelenterate, and some 

 Phit-worms.' Now tiie efficient arrangement of tlie pre-cellular 

 unit that gave rise to the cell was obtained by the massing of 

 chromidia (the essential chemical factor of the Organism) to the 

 centre, and by spreading around this central mass radiating fibrous 

 elements. The same principle is seen in the gastrsea, but here the 

 massed nuclei surround a central digestive cavity, while in the outer 

 zone the fibrous elements are specially developed and already are 

 beginning to difi'erentiate into muscles and nerves. The variety of 

 ccelenterate forms shows the passage in this period from the homo- 

 geneous to the heterogeneous, while colonies of Corals, Hydrozoa, etc., 

 are homogeneous aggregates. Bernard supposes that the next new 

 xinit was formed by a colony arranged so as to form a chain of 

 gastraeae with the Coeleutera of all the individuals connected up and 

 the muscular and nervous systems better developed. Such a colony 

 would be primitive annelid, and those annelids that branch and form 

 chains of individuals represent aggregates of the new unit.^ If the 

 passage from the gastrteal to the annelid period is scantily and 

 uuconvincingly demonstrated, still more so is that from the annelid 

 to the vertebrate period. Bernard dismisses this with very few 

 words ; and here is, perhaps, the weakest part of his book. A 

 scheme of animal evolution to be convincing should above all help 

 to elucidate the vexed question of the origin of the vertebrates, 

 instead of slighting it. After the vertebrate unit is reached, the 

 aggregates are social, and, since the tie connecting the individuals 

 is no longer a physical one, the scheme passes out of the natural 

 sciences into the metaphysical, and consequently no longer concerns 

 us as palaeontologists. 



III. 



It is not the object of this paper to give all the evidence for and 

 against the views put forward, and the reader must refer to the work 

 itself to form a true judgment. Had Bernard lived longer, no doubt 

 he would have tested his theories fully; as it is, he appeals to 

 liis readers to do so, and, if this sketch has been the means of 

 bringing his views to the notice of palaeontologists, it will have 

 accomplished an object. But the paper is primarily intended to show 

 how Bernard's theories may be applied directly in palaeontology and 

 give a helpful view of phenomena. 



Two outstanding points in Bernard's book are his insistence on the 



^ In embracing Mollusca among simple cell aggregates and apart from 

 annelids, surely Bernard fails to account for such forms as show traces of 

 segmentation and other suggestions of alliance with annelid worms, such 

 as the trochosphere larva ; see Pelseneer, " La Classification generale des 

 Mollusques " : Bulletin Scientifique de la France et de la Belgique, vol. xxiv 

 (1892), pp. 368-71, 1893, and " Recherches morphologiques et phylogenetiques 

 sur les Mollusques archaiques " : Memoires Couronnes I'Acad. roy. belgique, 

 vol. Ivii, pp. 83-8, 1899. 



' The position assigned by Bernard to Sagitta is interesting. It is an 

 aggregate of gastrseal units arranged on a different plan from the typical 

 annelid grouping and near that of the hypothetical vertebrate ancestor, but, 

 being of insufficient efficiency, it has failed in its turn to become a new unit. 



