THE LAST TEN YEAES. 361 



Denn wer die meisten Gestalten der vielfach umwolmeten Erde, 

 Die er vergleichend ersah, tragt im bewegenden Sinn, 



Wem sie die gliihende Brast mifc der fruchtbarsten Fiille durchwirken, 

 Der hat des Lebens Quell tiefer nnd voller geschopft. 



It will doubtless be remembered that the 'Survey of Nature,' 

 which forms the major part of the first two volumes of Cosmos,' 

 was described by Humboldt to Varnhagen in 1834 as 'the most 

 important part of the work.' The specific portion was intended 

 at that time to occupy but a subordinate position, serving 

 merely to add completeness without reverting to all the facts 

 previously mentioned, or entering upon any subject already 

 discussed. To this resolution, however, the author did not 

 adhere, as the material continued to grow under his hand ; 

 the empiricism of the age is everywhere apparent, while of the 

 graces of composition prominent in the 'Survey of Nature' 

 there is scarcely a trace ; the interest of the subject is para- 

 mount, and throws into the shade mere forms of expression. On 

 this account the later volumes present little that is peculiarly 

 characteristic of Humboldt: the 'Cosmos' of 1845-7 could 

 only have been conceived and written by Alexander von Hum- 

 boldt ; the plan of a specific description of the physical uni- 

 verse, as carried out in the third volume, might have originated 

 with any of our modern men of science, and in its execution 

 not merely one or two, but the whole of the present generation 

 of scientific men have taken part. It would not, in fact, be 

 too much to say that the specific portions of ' Cosmos ' are to 

 the ' Survey of Nature ' what the Pandects are to the Institutes 

 in Corpus Juris ; they are the incomplete pandects of science, so 

 far as science had been developed in the middle of the nine- 

 teenth century, and, like that digest of laws, are composed 

 merely of isolated data on the subjects in question gathered 

 from various authorities and put together after the fashion 

 of a mosaic. Humboldt shows himself therein as perhaps the 

 greatest compiler that has ever existed, distinguished alike for 

 industry, caution, accuracy, and comprehensiveness. A few 

 instances will suffice to justify these remarks. 



Even in the least speculative parts of ' Cosmos ' Humboldt 

 was never oblivious of the gradual development of science ; in 

 the notes, especially in those to the later volumes, much space 



