March 4, 1880] 



NATURE 



429 



in 1S34 he visited Great Britain for the first time, and his studies 

 received a fresh impetus from the wealth of new forms which he 

 found in English collections. In Scotland, too, collectors had 

 been bestirring them-elves, for besides what we have already 

 noticed as having been done by Sedgwick, Murchison, Hibbert, 

 and the Royal Society of K inburgh, Traill had made a valuable 

 collection from the Old Red Sandstone of Orkney ; Knight of 

 Aberdeen from the same formation at Gamrie ; Lord Greenock 

 had discovered the richness, in fish remains, of the Carboniferous 

 shales at W'ardie ; and many Scottish specimens had also been 

 collected by Jameson, Torrie, Buckland, and others. 



The British Association met in 1S34 at Edinburgh, and 

 Has then introduced by Buckland to the ( 

 immediately after Hibbert had read a paper, in which he 

 considered the gigantic teeth and bones found at Burdiehouse to 

 "resemble those of Saurian reptiles." Their piscine nature 

 ever, at once detected by the accomplished Swiss natu- 

 ralist, and the requisite material having been willingly handed 

 over to bim, he prepared and read, two days afterwards, a 

 "Report on the Fossil Fi-hes of Scotland," in which several 

 new genera are named. M 5! of the Scottish material obtained 

 by Agassiz at this time was | ublished in detail in the fasciculus 

 of his great work, which '^35, the Devonian forms 



including the genera Cepha/aspis, Aeanthoda, Chdracant/ius, 

 CheiroUpis, Diptertis, and Osteolepis ; while those from Carboni- 

 cks were referred to Amblyptents, PaUetmiscus, Bury- 

 ttotus, Pygepterus, Megalichthys, Gyracanthus, Tristichius, 

 C ten of/} <ck ius, &c . 



Agassiz revisited Scotland in 1842, and was pre-ent at the 

 meeting of the British Association held that year at 



le the material for the further study and description of 

 Scottish fossil fib remains had vastly increased. Large collec- 

 tions from the Old Red Sand tone beds of Cromarty and Moray- 

 shire had been made by Ilu.'h Miller, Dr. Malcolmson, Lady 

 mnming, and Mr. Alexander Robertson. The collec- 

 ' ord Enniskillen .and Sir Philip Egerton, which already, 

 at the time of Agassiz's 



species, now con- 

 tained a choice .-election S from Si coniferous 

 I been assiduously c Heeled by Dr. Rankin of Carluke 

 and others. The large acce sion of material from the Old Red 

 enabled Agassiz in 1S42 to lay before 1 

 n a "Report on the Fossil Fishes of the Devonian 

 System," which finishes with a list of fifty-live species belonging 

 to twenty genera. 



His great work, the " Recherches sur Ies Poi 



: ted in 1S43, a '"l '" it was inserted a general list of 

 ■11 the fossil fishes which had till then come under his notice. 

 Here we find ninety-nine species named from Scottish 

 but, unf irtunately, descriptions only of twenty-five were included 

 in the text. The others he re ervei i t ir a projected serie 

 menlary monographs, of v. bich only one ever a | 

 that on the fishes of the Old Red Sandstone, which 

 pleted in 1S46. In this work sixty-seven Scottish 



1 described, and some improvements in classification 

 the establishment of the new families of Cephaldspida, 

 ■ ■■:', the two former being dis- 

 from the old heterogeneous Lepidoidd, and the latter 

 111 the Lepidoidd and partly from the so-called 



In offering a few words of comment upon the labours of 

 di partment, the highest tribute of honour must 

 be paid to him for the position to which he rai ed the science of 

 the enormous amount of work 

 which he accomplished in so short a time. Eminent as well in 

 other branches of zoology, his name w ill gj down to posterity as 

 that of one of the greatest naturalistsof the present century. To 

 him we owe the establishment of the order of Ganoid fishes,- the 

 description of an enormous array of genera and species, and the 

 first valuable generalisations as'to the history and sur 

 ichth) ic life on the globe. An opponent of the so-called vertebral 

 theory of the skull, as held by Oken/and modified by Owen and 

 Others, as well as of the doctrine of descent, he nevertheless 

 pointed out what, a; Prof. Marsh says, "is now thought to be 

 Strongest points in favour of evolution," namely, the 

 correspondence between the hcterocercal character of the tail in 

 the embryos of modern osseous fishes, and the prevalence of that 

 * jim among the adult fishes of the older formations, stating, in 

 tact, that " les poi-sons fossiles du vieux gres rouge represented 

 raellement l'age embryonique du regne des poissons." But it Ls 



hardly possible for the zoologist of the present day to suppress 

 some feeling of wonder that a man, so well versed in general 

 zoology and anatomy as Agassiz, should have based his classifi- 

 cation of fi-hes upon characters so trivial as the mere external 

 aspect of their scales, or that he should have distinguished many 

 of the families into which he divided the order of Ganoids by 

 characters equally superficial. We may quote, for instance, his 

 inclusion among the Ganoids of the Pipe-fishes, Siluroids, Globe- 

 fishes, and Trunk-fishes, merely on account of their bony scutes ; 

 the entirely artificial nature of the distinction which he drew 

 between his Ganoid families of " Lepidoids " and "Sauroids," 

 and the consequent utterly heterogeneous character of both ; the 

 similarly unsatisfactory nature of his family of Caiacant/ii, into 

 which he even introduced the recent Teleostean Arapaima; — 

 and so on. However, it is at the same time only natural that he 

 should have been imperfectly acquainted with the anatomy of 



i Is, considering the as yet comparativi 

 material at his disposal, and it is also evident that, had he 

 ae to the elucidation of Osteological detail, he 

 have gone over the same enormous amount of 

 ground within s 1 limited a period. 



1 of fishes was at first eagerly accepted 

 by geologists and callers, largely on account of its 



e. It c mid not, however, stand the test of anatomical 

 inquiry, and was soon superseded by the system proposed by 

 Johannes Miiller in 1844, which, with various minor modifica- 

 tions, is the one still adhered to by most zoologists. Such, 

 however, was the influence of Agassiz, and such the supposed 

 ff convenience" of his system, that we find it in use, especially 

 amongst g " palaeontologists," years after Midler's 



great paper " l el cr den Bau und die Grenzen der Ganoiden" 

 was published. 



The large fossil creature whose laniavy (eetb, sometimes four 

 or five inches in length, suggested the idea of a "Saurian 

 reptile" to Hibbert, and which was rightly placed among the 

 fishes by Agastiz, received from him the not inapproprial 



Hibbarti. With its remains, however, those of 

 taller fish, with glossy angular scales, were at the time 

 unfortunately confounded, but there can be no doubt that the 

 lichthys vrtLS suggested by the large teeth, and pro- 

 perly belonged to their possessor. Nevertheless, some time 

 , on visiting Leeds, and finding in the Museum there 

 the head of an example of the smaller fish, Agassiz described 

 i it in a subsequent number of the "Poissons fossiles" 

 as Megalichthys Hibberti, while for the real and origin il 

 hchthys, along with some Old Red species he founded the genus 

 Hehptychius. Prof. Owen, however, in his "Odontography" 

 (1S40-45), elevated the Carboniferous " Lloloplychins" 

 into the new 'tis, giving also many important details 



structure of the teeth. The claims of 

 ric distinction were stoutly disputed by 

 in his work on the fi-hes of the Old Red Sandstone. Subsequent 

 investigation has, however, not only proved the validity of 

 ius, but also that it cannot even be in 

 family with Jloioplyehius. In the same work Owen 

 described the remarkable microscopic structure of tl 

 teeth from the Old Red Sandstone of Morayshire, to which he 

 gave the name of Dcndrodus. 



The next writer on Scottish fossil fishes who claims attention 



IS Hugh Miller, who devoted his chief attention to them, and 



whose collection of "Old Red" forms furnished many of the 



iribed and figured by Agassiz in his "Mom 



a Vieux Gres Rouge," as well as man; 

 which were also figured by himself. 



Among Miller's fascinating popular descriptions of scenery, 



geological structure, and fossil fishes, we find some genuine 



touches of original palteontological observation, which quite 



indicate what his powers in that direction might Lave 



We find, for instance, 



ts quite au are that Cheirolefis was not an Acanthodian, 



though it was classed by Agassiz in that family. We find a 



very creditable restoration of Osttolepis, infinitely superior to 



that given by Agassiz some years afterwards, and hardly inferior 



iven by the accomplished Pander; and we find him 



correctly interpreting as the ventral surface of Pterichthys that 



aspect of the creature erroneously represented by Agassiz as the 



dorsal. lie also showed that Ag ractus, sup]. 1 ed 



by him to be a genus allied to Pterichthys, was nothing more 



than the cranial shield of a Dip'.erus. He likewise discovered 



the dentition of JJiptotis, which, with the structure of the palatal 



any 



