380 On the Genus Ortmannia, Rathb. 



In face of these modifications, one cannot doubt but that the 

 Atyce are the direct descendants of Ortmannia, and that, in the 

 case of certain species, this derivation is not yet a definitely 

 accomplished fact. It is naturally among the^ small forms, 

 nearer than any others to the primitive Atyidae, that this 

 condition of unstable equilibrium is seen still to exist, in 

 which the same creature may indifferently present the form 

 of the past or of the future : Atya Usulcata and Atya serrata 

 are still in this stage. In Ortmannia americana the 

 primitive form alone exists ; either it has persisted after 

 having produced the Atya, or it is in a state of evolution 

 towards the production of this kind, which is more probable. 

 In Atya brevifrons, de Man, on the contrary, the primitive 

 form seems to have disappeared, bequeathing a very marked 

 stamp to its descendant, which is small like the Ortmannia 

 and provided as it is with locomotor feet of small power. 

 A. brevifrons is a common species in the islands of the 

 Pacific ; it has never been noticed under the form Ortmannia, 

 but it is possible that in some island it persists still in that 

 state. 



It goes without saying that in the most typical Atyo2 

 {A. robusta, A. scabra, &c), which are greatly modified and 

 of large size, one would not expect to find specimens having 

 the Ortmannia form. 



Here, then, manifestly are mutations by atavism which 

 show us how new types are formed and old types persist. 

 Actually, Atya bisuhata and A. serrata are represented by 

 individuals of two kinds — the one with chela? split down to 

 the base, the other with normal chela}. If these species were 

 social, the individuals of each type might be called upon to 

 play a different role in the colony, and to a certainty the 

 characters which distinguish them would go on exaggerating 

 themselves in consequence. 



May we not explain in the same way the mysterious 

 presence of polymorphic individuals in the societies of ants 

 and termites? and the starting-point of the polymorphism 

 of these forms, w r ould it not be an atavic mutation similar 

 to that of the Atyce? 



1 return to the domain of pure systematics. The genus 

 Ortmannia should persist, but it comprises up to the present 

 time, it appears, only a single independent species — 0. mexi- 

 cana, of Tropical America. The modification Henshawi of 

 Atya bisulcata and the modification Alluaudi of Atya serrata 

 are clearly Ortmannia; but they represent species in course of 

 evolution, which, according to circumstances, may persist or 

 disappear as Ortmannia ; it is useful to look upon them no 

 longer as independent species, but as the atavic form of the 



