18 



should not call artificial divisions b)' this name. The characters of artificial genera depend 

 solely upon the taste of the worker and the convenience of separating into groups animals 

 and plants. All species are considered to belong to the same natural genus which agree 

 in structural characters, external and internal, or anatomical ones in the different stages, 

 in transformation, in the manner of living. These definitions of a genus are accepted 

 as well by naturalists who are strong Darwinians as those who oppose the develop- 

 ment theory. In a prize essay of the Jena University, D. P. Mayer, a pupil of Prof. 

 Haeckel, in a paper on the " Ontogeny and Phylogeny of Insects," enlarges this definition 

 in so far as he asks for a conformity in the embryological characters. I believe no one 

 will object that this definition is a good and exhaustive one ; but if we attempt to use it 

 in a special case we become bewildered by the astonishing amount of characters unknown 

 to us, and the impossibility to make them out for our work. At present we know hardly 

 well enough the external character of the imago. Of other characters our knowledge is 

 merely fragmentary and often a tabula rasa. We may say that a century of hard work 

 will not fill these gaps in our knowledge. It is obvious that we cannot wait till this 

 enormous amount of work is done. And it is certain that naturalists will not and can not 

 stop creating new genera. 



Genera created with such a limited amount of knowledge will depend upon the ex- 

 perience and taste of the worker. Many of such genera will have to be modified or 

 dropped by a farther advancing knowledge. 



The most important question (what are generic characters'!) is still unanswered. 

 The large literature and the difference of opinion emitted by prominent authorities 

 seem to prove that a sufficient affirmative answer is impossible till our knowledge is fur- 

 ther advanced. But here, as in other abstract questions, we can proceed in a negative 

 manner by exclusion. 



Genera consist of a number of related species. If we knew the character of the 

 species, the specific character, we can by exclusion come nearer the character of the genus. 

 Species differ by structural character, and as the species form the lowest degree of the 

 classification, we can be sure that species must differ at least by minutest points of 

 structure. 



I think there is no objection of consequence possible. I know very well that differ- 

 ences in minuter points of structure have been considered as generic characters. But 

 naturalists beginning with the construction and definition of the higher degrees of class, 

 order, family, &c., used up all characters at hand, till, coming to genera, nothing was left 

 but minute differences of structure ; the simple consequence of using specific characters 

 for generic ones was that nearly every species was considered to be a genus. 



I said before that species must differ at least by minuter points of structure. The 

 discovery which I mentioned before proves that structural characters ot species are more 

 important, and can by a different manner of living be changed in such a way as to repre- 

 sent forms which were formerly believed to belong to different genera. Branchipus and 

 Artemia, belonging to the Phyllopod Crustacea, are represented by several species here 

 and in Europe. The two genera are nearly related one to the other, and differ principal!}' 

 in the following points : Artemia has eight post-abdominal segments, the last one very 

 long. Branchipus has nine post-abdominal segments, the last two of equal size. Artemia 

 has three articulated claspers in the male ; Branchipus two articulated claspers. Artemia 

 is often propagated by Parthenogenesis, Branchipus never. 



Nobody will deny that those characters of structure go very far beyond minuter 

 points of structure, and are marked well enough to justify the separation sixty years ago 

 by Dr. Leach. Now it is proved that not only the species of Artemia known up to-day 

 from Europe, Asia and Africa, but even some species of Branchipus belong to one and the 

 same genus and species. In the American fauna five species of Artemia and three of 

 Branchipus are described ; of course they will have to be studied again in a similar manner 

 as the European ones. The twoEuropean species oi Artemia are remarkably different. Artemia 

 salina has a strongly bifid tail surrounded by 15 to 20 bristles and narrow gills; sirtemia 

 miilhavscni has a rounded tail without bristles and very large gills. This latter species 

 lives in pools of a very concentrated salt water of 25° Beaum6 ; the other species in common 

 salt water of about 8°. In 187 1, a dam which surrounded a salt pool containing Artemia 

 mulhauseni, broke down by accident, and the sea water washed in at the same time ; Ar- 



