OBSERVATION AS AGAINST UNDUE GENERALIZATION. I063 



This generalization was applied to the cichlids of Africa and Palestine and, 

 in various accounts of the habits of the bolti and similar fishes, reputable writers, 

 such as Giinther and Lortet, especially credited the males with exclusive parental 

 care. Subsequent dissections of the same species and other species observed 

 by these naturalists revealed the fact that in all the cases in question the females 

 took charge, taking their eggs in their mouths and caring for them and the 

 newly hatched young until they had attained a considerable size. In fact, so 

 far as the cichlids are concerned, numerous African species have now been 

 examined, and for all of those so examined the females have been ascertained to 

 be the egg-carriers. Let it not be assumed, however, that all the other cichlids 

 take such care of the young and that all American species do so, as well as the 

 African. Indeed, even now it is known that certain South American species 

 provide for their eggs in nests made by heaping pebbles over their eggs or other- 

 wise preparing the bottom, rather than by oral incubation. But more than 

 this is not known and we are ignorant of the parts played by the respective 

 sexes. 



The tendency to undue generalization has been exhibited in a striking and 

 even amusing manner in the case of two European fishes already referred to, 

 the wels of Germany arid the glanis of Greece. The wels had long ago been 

 declared by many observers to exercise no parental care after deposition and 

 fertilization of their eggs. It happened, however, that Aristotle, over twenty- 

 two centuries ago, gave elaborate details of the glanis and the care taken of 

 the eggs by the male parent. Instead of those accounts, which bore the impress 

 of observation and truth on their face, serving as a check to identification, it 

 was assumed by some of the greatest of modern ichthyologists, such as Cuvier, 

 Valenciennes, and F. A. Smitt, that the wels and the glanis were of the same 

 species; the Frenchmen declared that Aristotle's account "borders a little on 

 the marvelous ' ' and the Swede reechoed with the remark that "it is now 

 regarded as dubious." Yet over half a century ago (1856) Agassiz declared that 

 Aristotle was right and that the Aristotelian fish differed, not only specifically 

 but generically, from the wels. Later, comparative descriptions and illustra- 

 tions of the Grecian species were published; nevertheless the two continued to 

 be confounded in Europe under the same name. But recently a new attitude 

 has been assumed otherwise. At last it has not only been acknowledged that 

 the facts recorded by Aristotle were credible, but assumed that what was true 

 of the glanis must be true of the wels; in a reference to the species made by a 

 distinguished French ichthyologist this year the wels (' ' I'enorme Silure d' 

 Europe ' ') is credited with paternal instinct and attention. Thus has generaliza- 

 tion been carried to an extreme and assumption piled on assumption. One 

 further assumption apparently was that because the glanis was not in a European 



