328 NATURAL SCIENCE [November- 



capable of appreciating the fossil and preserving it for study when 

 discovered. 



The importance of remembering these considerations when 

 speculating on biological subjects has recently been illustrated once 

 more by the discovery of a new Upper Silurian fish-fauna in the 

 south of Scotland. As already mentioned in Natural Science 

 (vol. xiii., p. 157) this remarkable assemblage of fishes or fish-like 

 organisms has been found by the Geological Survey at the top of 

 the Silurian formations of Lanarkshire ; and some preliminary 

 notes by Dr Traquair announce that a complete memoir on the 

 subject will shortly appear. Now, scattered and abraded fragments 

 of similar organisms have been known for nearly sixty years in a 

 thin stratum, termed the Ludlow Bone-bed, in the Upper Silurian of 

 Herefordshire and adjoining counties. Fossils of the same kind 

 have been collected for nearly half a century in an Upper Silurian 

 limestone in the island of Oesel, in the Baltic Sea. Traces of them 

 also occur in Galicia, Pennsylvania, and New Brunswick ; and a 

 few years ago similar fragments were sent to me by the Geological 

 Survey of Canada from another locality in Newfoundland. How- 

 ever, notwithstanding this proof of the very wide distribution of the 

 late Silurian fish-fauna in question, we have had to wait for the 

 accidental discovery of a thin stratum in Lanarkshire to obtain even 

 a faint idea of the strange types of life represented by the familiar 

 scales and other exoskeletal fragments. 



Although it is now nearly forty years since Darwin's " Origin 

 of Species " first appeared, his lament at the hopelessness of testing 

 all the principles of organic evolution by reference to the " records 

 ■of the rocks " might indeed be appropriately renewed at the present 

 day. The discovery of new fossils in all parts of the world has pro- 

 gressed at an astounding rate in the interval; and we are beginning 

 to perceive feebly some of the laws which govern their succession 

 and distribution. The biologist who is prone to glance through 

 palaeontological text-books, however, and utilise them in his specula- 

 tions, cannot be too frequently warned of the imperfection of our 

 knowledge and the danger of trusting to negative evidence. 



To understand the importance of this warning at the end of 

 nineteenth-century science, it is only necessary to consider the case 

 of some of the most striking and philosophically valuable vertebrate 

 animals. 



Firstly, there is the remarkably early Devonian organism 

 Palacospondylus gunni, frequently referred to in these pages. 

 Whether it is a primaeval lamprey or not, it is the single known 

 representative of its group, and implies the former existence of a 

 great race of which we are acquainted with no other member. This 

 fossil occurs in the Caithness flagstones, which were deposited in a 



