256 The Edinburgh New Philosophical Journal. 



we shall only remark, that nearly all authors have shared in the 

 ideas of Leuwenhoek ; and that it was impossible for them to com- 

 prehend truly either the form or organization of scales, from not 

 having employed the most powerful means of investigation, that is 

 to say the microscope, which has procured more complete results 

 for us.* 



Chap. III. — Scales considered as affordiny Characters for Classification. 



"The most important question in the study of natural history, is 

 the determination of natural characters, that is to say, of characters 

 which bring together the inherent properties of animals, — which are 

 proper to them in their natural state, — and which can be studied either 

 by observing an entire body, or only a few detached parts, and that 

 always without the assistance of other sciences. Now, what is most 

 important in these researches, is precisely to determine characters in 

 intimate relation with the nature of the animal, its organization, the 

 anatomy and physiology of its body, which are, in short, essential, 

 and undergoing no change with the accidents of habitation, food, &c. 



"It is the just determination and appreciation of these characters 

 which can furnish us with true and precise notions of species, genera, 

 and families, and enable us to avoid the errors committed by those 

 who imagine they have found, in the most insignificant details, sufficient 

 characters to constitute a new species. Have we not thus seen botan- 

 ists make two different species of the same flower, as it happened 

 to grow on the sides or in the bottom of a ditch, and might, from 

 this cause, more or less differ in the brilliancy of its flower ? Do we not 

 daily see ichthyologists and zoologists creating new species according to 

 changes of colour, size, &c, — changes which are altogether accidental, 

 and depend entirely on the influences of climate, food, &c. But, apart 

 from the small value attaching to such characters as these, there is 

 still another point most commonly overlooked by naturalists, on which 

 we wish to say a few words. 



" Any property whatever, which constitutes a natural character, may 

 vary in different degrees, and thus constitute in its successive changes 

 a continuous series. Relations unite all the members of this series ; 

 no one is separated from another by a real difference, and each mem- 

 ber of this series may be replaced by another of the same series, 

 without changing the nature of the being. Thus, for example, white 



" * The second chapter of Dr. Mandl's paper is devoted to the explanation of the structure 

 of scales. This we have been obliged, in the mean while, to omit, as it requires a greater 

 number of illustrative figures to render it fully understood, than we can at present find room for. 

 The omission, however, is less to be regretted, as it in noway affects the interest or value of the 

 two other chapters, each of which is complete in itself. — Editor. Edik. Phil. Jour. 



