IRREVERSIBLE EVOLUTION PETRONIEVICS. 435 



Another important instance of the etiological application is fur- 

 nished by the bipedal habits of the immediate ancestors of Stego- 

 saurus and Triceratops. 



If evolution were reversible these two dinosaurs would have exactly regained 

 their former quadrupedal structure, and there woidd have been no way to dis- 

 tinguish their secondary quadrupedal existence from the first (Dollo, 10, p. 448). 



The other most important cases of etiological application are : The 

 secondary adaptation to the swimming sea life of the Pycnodonts 

 (Dollo, IT, pp. 108-9), the secondary adaptation to the swimming sea 

 life of the Trilobites Dephon and Aeglina (Dollo 16, pp. 410 and 

 412), etc. 



Among the instances of the morphological application of the law, 

 that of the secondary abdominal ventral fins in the teleosts has a spe- 

 cial importance. As is known, the ventrals of teleosts may be either 

 abdominal or thoracic or jugular. But among the abdominal ventrals 

 we have two types — those which have no connection whatever with 

 the pectoral girdle, and those joined to the clavicular symphysis by 

 a ligament. As there is no reason for the presence of this ligament 

 in situ we have to conclude that it is the degenerate remnant of a 

 former direct connection with the pectoral girdle. In conformity 

 with the irreversibility of evolution the ventrals in again becoming 

 abdominal have kept the connection with the clavicular symphysis 

 which they acquired when occupying a thoracic or jugular position 

 (Dollo, 14, p. 139). 



The other important instances of the morphological application 

 of the law are: (1) The very anteriorly placed choanae of the sea 

 turtles (Dollo, 8, pp. 817-820), (2) the longirostral and brevirostral 

 condition in Crocodilians (Dollo, 12, p. 85), etc. 



Finally, we must mention the one instance in which Dollo has 

 used his law in systematic work — the Ptyctodonts. Before Dollo 

 these fossil fishes, then known from their dental plates only, had been 

 placed among the Holocephali. In his important memoir on this 

 subject (Dollo, 13) Dollo showed that, by virtue of the law of ir- 

 reversible evolution, the Ptyctodonts can not be regarded as Holo- 

 cephali and that they ought to be treated as Arthroderes. Since 

 then Dollo's conclusion has been wholly confirmed. 



Although the empiric evidence for the validity of his law has been 

 abundant and varied, Dollo was not satisfied with such a wholly 

 empiric demonstration. He has attempted to give a deductive dem- 

 onstration as well. He says : 



The Irreversibility of Evolution is not, as many have believed it to be, 

 merely an empiric law based purely on facts of observation. But it has deep- 

 seated causes which carry it in final analysis to a question of probabilities, 

 as in the case of the other laws of nature. Evolution being a summation of 

 exactly determined individual variations in an exactly determined order, to 

 have it reversible would be to admit the possibilty of the intervention of 

 causes exactly the inverse of those which produced and fixed the individual 

 variations from which the first transformation arose, and in an exactly inverse 



