Agassis 



Be sure that the methods pursued predetermine 

 the results. It is as important to be careful of 

 methods aa of facts. Every chemist knows this. 

 Chemistry and Physics are far in advance of Botany 

 and Zoology as exact sciences. The chemists and 

 physicists test their methods with more care than 

 do naturalists. The astronomer goes further still, 

 and takes his own physical organization into ac- 

 count. Ho applies a correction to his observation, 

 which is known as the "personal equation." The per- 

 sonal equation should be allowed for by naturalists. 



Linnaaus, who had an eye for affinities such as few 

 men ever had, was the first to introduce some con- 

 siderations on the classification of fishes, which have 

 had a permanent influence. He separated fishes 

 like sharks and skates, with cartilaginous skeletons, 

 from the bony fishes. Then he divided the bony 

 fishes according to the position and characters of the 

 fins. Astedi opposed this classification, and though 

 at first unheeded, Cuvier afterward adopted his 

 idea, which was that bony fishes should be separated 

 into classes Acantiiopterygia, with sharp spines in 

 the dorsal fins, and Malacapterygia, fins without 

 spines. 



When I began to study, I wished to iden tify fossil 

 Sshes with those now living. But after studying 

 fossil fishes for many 3 r ears, I found that the old 

 classification, based upon cartilaginous and bony 

 skeletons, would not answer. I tried the scales to 

 see what basis they would otter for classification. 

 1 divided the scales, according to their form and 

 structure, into four kinds, to which I applied the 

 names, Cycloid, Ctenoid, Ganoid, and Placoid. 

 When this classification was compared with that of 

 Artedi, it was found, curiously enough, that the Cy- 

 cloids and Cteuoids included almost exactly the 

 same species as the Acauthopterygiaus and Malacop- 

 terygians. Fishes like the sturgeon and garpike, in- 

 cluding all fossil fish from the Siluriau up to the 

 Cretaceous, not counting sharks. have Ganoid scales. 

 Sharks and skates have Placoid scales. 



The sturgeon was the battle-field of naturalists 

 for 10 years, but the question is settled now, and tho 

 sturgeon is classed with the garpikes and other Ga- 

 coids. Johannes Miiller found no two fishes more 

 closely allied in anatomical structure than garpikes 

 and sturgeons, which I had already shown to be 

 closely allied with respect to their scales. The gar- 

 pikes have reptilian characteristics. 



We see the importance of grouping animals by 

 their general resemblances, criticising every step as 

 T*e go< _,Do not not take up definite features as dis- 

 tinctive ctaracters, but take the whole animal and 

 classify by a general resemblance of all tho features. 



Johannes Miiller, the master of Comparative 

 Anatomy at present, has made divisions resting 

 upon distinctions which are physiological and Dot 



at 



zoological, considering tinners of paramount impo* 

 tance which ought not to be so considered. Tho 

 presence or absence of the nictitating membrane In 

 makes the basis of a distinction between two classed 

 of sharks, while in reality every vertebrate at a cer 

 tain stage of growth has a nictitating membrane. He 

 divides selachians into two classes, according aa 

 they have or have not spiracles, when really there 

 is a perfect gradation from those which have nous 

 to those which have perfectly developed spiracles ; 

 thus widely separating animals which are closely re 

 lated in structure. 



Embryology will lead to better things ; but we 

 should not follow the indications of embryology 

 implicitly, but question every step. It will show ua 

 differences akin to the difference among adult ani- 

 mals. In the egg the structure is simple. The p;o- 

 gress in structural complication is consecutive. We 

 must examine the character of our standard as well 

 as of our facts. 



The classification of mammals, though further ad- 

 vanced than that of birds, is still not based on the 

 true affinities of structural complication. Th 

 manatus, or sea-cow, for instance, which is usually 

 classed with the whales, is shown by the study of 

 fossils to be closely allied to the elephant. 

 Batrachians are closely related to ono 

 another anatomically. The lowest ba- 

 trachians resemble the embryonic forma oi 

 the higher batrachians. Among true rep- 

 tiles, no embryonic affinities can be traced as in 

 batrachians, The embryo turtle does not resemble 

 any other reptile. At no period in its life is itsnak 

 like or lizard-like; but at a very early stage it rG'? 

 sembles birds. It points forward, not backward. IS 

 is on anatomical grounds, structural complication, 

 that we place the lizard before the snake and the 

 turtle before the lizard. The crocodiles stand before 

 ordinary lizards in complication of structure. They 

 also have characteristics that ally them to the ex- 

 tinct fauna of past geological ages. Though snakes 

 are later than the geologic representatives of croco- 

 diles, crocodiles are higher than snakes. 



A synthetic type is a group of animals in which 

 characters found independently in other groups are 

 found combined. Lizards are a synthetic type. 

 They are comparatively old in geologic time, and 

 these geologic forms are not among the simplest, for 

 the first reptiles were higher than any that ever fol- 

 lowed. Make transmutation of this if you can. It 

 is not true that animals have followed each other in 

 successive progress and development. 



Prophetic types are animals that at an early period 

 combine characters which are found in animals of e 

 higher order in later time. We have among these 

 natural series from lower to higher structures. Tbr 

 gradation is a structural gradation, and in the em- 



