492 



SCIENCE 



[N. S. Vol. XXVI. No. 668 



of Charleston, S. C, who mistook the pe- 

 culiar transversely expanded and partly 

 double air-bladder of Diodon for a lung. 

 With such errors of observation as a basis, 

 Linne apparently assumed that all the as- 

 sociated genera also had lungs. Gmelin, 

 in his edition of the "Systema Naturte" 

 (generally called the thirteenth), corrected 

 this error, and returned all the Nantes to 

 the class of Pisces, thus reverting to the 

 older view of Linne himself. The Pisces 

 of Linne included only the genera left 

 after the exclusion of those just named 

 and also of Myxine, which last was re- 

 ferred to the class of Vermes between the 

 leeches (Hirudo) and the ship-worms 

 (Teredo). 



LINN^AN GENERA 



The genera of Linne were intended and 

 thought by him to be natural,' and natural 

 groups some of the so-called genera were, 

 but present opinion assigns to most of them 

 a very different valuation from that given 

 in the "Systema Naturse." Some of the 

 genera of invertebrates were extremely 

 comprehensive. For example, Asterias in- 

 cluded all the members of the modern 

 ■classes of Stelleroidea or Asteroidea and 

 Ophiuroidea ; Echinus was coequal with 

 the Eehinoidea; Cancer, Scorpio, Aranea, 

 Scolopendra and Julus were essentially 

 coextensive with orders or even higher 

 groups of the zoologists of the present time. 

 Others were so heterogeneous that they can 

 not be compared with modern groups. 

 Thus Eolothuria, in the last edition of the 

 ' ' Systema, ' ' was made to include four holo- 

 thurians in the modern sense, a worm, a 

 Physaliid, and three tunicates; in other 

 terms, the so-called genus included repre- 

 sentatives of four different classes, and 

 even branches of the animal kingdom. 



It has been stated by various writers 

 that the genera of Linne were essentially 



' Classis et ordo est sapientife, genus et species 

 Natures opus. — Linn. " Syst. Nat.," I., 13. 



coequal with the families of modern au- 

 thors, but, as has been indicated, such is 

 by no means the case. Other striking ex- 

 ceptions to the generalization may bfe 

 shown. 



Not a few of the genera of Vertebrates, 

 although not of the superlative rank as 

 several of the Invertebrates, were equiva- 

 lent to orders of modern zoology; such 

 were, in the main, Simia, Testndo, Vesper- 

 tilio and Rana. Simia included all the 

 anthropoid Primates or monkeys except 

 man; Vespertilio was equivalent to the 

 order Chiroptera less the genus Noctilio; 

 Testudo was exactly equal to the order 

 Testudinata or Chelonia; Rana to the or- 

 der Salientia or Anura. A number of 

 other genera of one or few species known 

 to Linne were also of ordinal or subordinal 

 value. 



In striking contrast with the range of 

 variation of such genera were others, of 

 which several, weU represented in northern 

 waters, may be taken as examples. Scor- 

 pcena was distinguished simply because it 

 had skinny tags on the head;* Lairus be- 

 cause it had free membranous extensions 

 behind the dorsal spines;^ and Cobitis be- 

 cause it had the caudal peduncle of regular 

 height" and scarcely constricted as usual in 

 fishes. These characters are of such slight 

 systematic importance that they have not 

 been used in the diagnoses of the genera 

 by modern ichthyologists. Further, use of 

 them misled even Linne as well as his suc- 

 cessors. Some of the consequences may be 

 noticed. 



The close afSnity of the "Norway had- 

 dock" or Swedish Kungsfisk or Rodfisk 

 (Seiastes marinus) to the typical Scor- 

 pcena was unperceived and that species 



' Scorpwna. Caput cirris adspersum. 

 ° Lairus. Pinna dorsalis ramento post spinas 

 notata. 



° Cobitis. Corpus vix ad caudam angustatum. 



