308 On the chief Generic Types of Palceozoic Corah. 



study. One course would be to throw all such forms together, as 

 constituting a single "genus/' and to consider the different types 

 of the series as so many " subgenera." The objections to this 

 course, however, are twofold. In the first place, as our know- 

 ledge increases, and as we obtain a more complete series of forms 

 uniting different groups together, our genera would undergo a 

 proportionate expansion and increase of size, until, in the case 

 of fossil forms at any rate, they would become utterly un- 

 manageable. A " genus " might in this way come ultimately 

 to include all the forms which now constitute a " family ;" and 

 from the number and variety of the forms included in it, there 

 would be great difficulty in giving it any precise definition. 

 In the second place, " subgenera," however appropriate and 

 justifiable in theory, are nearly useless, if not absolutely 

 obstructive, in actual practice. We think most working natu- 

 ralists will bear us out in this view, at any rate so far as con- 

 cerns all cases where we cannot obtain complete materials for 

 determination and examination, as we never can in the case of 

 fossils. 



On the other hand, the course may be adopted of selecting 

 certain well-marked and easily recognizable "types," and of 

 giving to these distinctive names. By adopting such a course 

 the actual practical work of determining and classifying spe- 

 cimens is enormously facilitated ; and it does not matter to the 

 observer, though his " types " be connected together by a gra- 

 dational series of transitional forms. He does not thereby 

 hamper his theoretical conceptions as to what constitutes a 

 " genus " or a " species ; " nor is he pledged to believe that 

 his selected types are all of the same value. He is not even 

 bound to express a positive opinion as to whether his " types " 

 are to be regarded as " genera " or as " subgenera." This, 

 indeed, must be a comparative matter of indifference to him, 

 since, upon all theories of Evolution, these terms are of merely 

 conventional value, elastic, and incapable of precise definition, 

 and owing their existence at all, in any definite form, to our 

 own imperfect knowledge- Nor does it seem that the reproach 

 can justly be made that the adoption of this course opens the 

 door to a wide and undesirable multiplication of names. 

 " Subgenera" have distinctive names applied to them just as 

 much as " genera," with the disadvantage that they are in- 

 finitely more perplexing and more difficult to work with. 



In working out the Palaeozoic corals we have thought it 

 best to adopt the course above indicated. We have chosen 

 certain structural " types," which the observer can always 

 recognize with certainty when lie is provided with properly 

 prepared specimens ; and to these we have given distinctive 



