REPORT OF THE DIRECTOR I919 



75 



possess " one common characteristic which seems to annul their 

 significance as evidence of reversion," namely, " they all relate to 

 meristic variations in which certain organs, of which at least one 

 already exists, suffer an increase in number" (as p. e. the stamens). 



According to Arber, the hypothesis of reversion has too often 

 been employed by morphologists in a noncritical spirit, and " the 

 only instances of genuine atavism of, which we have any 

 knowledge are those which consist in the synthesis by hybridization 

 of some original form which has now become split into different 

 races by loss of factors " 



Another cause of " atavism " or " reversion " is not infrequently 

 seen among fossils where a large number of species characterizing 

 a certain horizon (see Grabau, 1913, p. 974) will be found " in 

 which ancestral characters occur in the adult, thus recalling species 

 of a lower geologic horizon." " This atavism or reversion of the 

 species to ancestral characters miay often be seen to be nothing more 

 than an arrestation of development at an immature morphic stage, 

 when the characters of the young are like those of the adult ancestor 

 of earlier geologic horizons." 



In looking at our case of reversion in Triarthrus spinosus 

 in the light of the views just recorded, it is necessary to consider 

 separately the forward carrying of the genal spine and the pos- 

 terior series of long thoracic spines. 



It is obvious from the development of Elliptocephala 

 asaphoides and other Cambrian trilobites, that the earlier 



Fig. 15 



; '^ Fig. 17 



Fig. 18 



Fig. 19 



Fig. 20 



Figs. 15-17 Elliptocephala asaphoides Emmons. Figs. 18-20 

 Olenellus fremonti Walcott. Growth-stages showing the forward 

 position of the genal spines. (All after Walcott) 



