368 ALEXANDER VON HUMBOLDT. 



and as irrational ' as the French book of synonyms which had 

 once been c recommended by the Abbe Delisle.' l 



While in the compilation of the first two volumes the 

 outlines already existed in the notes of the lectures delivered 

 in 1827, there was wanting for the latter volumes even this 

 meagre guide. These were composed straight off, a sheet and 

 a half at a time, and were written on small pieces of paper, 

 whence they were transcribed for the press by Buschmann. 

 This complete reconstruction, c scarcely two months before 

 going to press,' indispensable in a work purporting to give 

 a c picture of the state of science in the middle of the 

 nineteenth century, embracing not only accepted views and 

 theories most deserving attention, but numerical data of all 

 kinds, given with the greatest attainable accuracy,' was un- 

 doubtedly the cause of that admirable freshness 2 which must 

 ever be a subject of wonder in a production of extreme old 

 age; it is, however, to this editorial process that is to be 

 ascribed the unfinished condition of the work, reminding one 

 of a torso. As the mortal remains of the distinguished author 

 were being conveyed to the Cathedral on May 10, 1859, there 

 arrived from Stuttgard the proof-sheets of the manuscript that 

 had been forwarded on, April 19 ; but, unfortunately, no clue 

 existed to guide another hand in completing the work from the 

 material collected. 3 



Nevertheless, though unfinished, not perfectly homogeneous, 

 displaying a want of adjustment between the claims of science 

 and the elegancies of literature, and now to some extent obso- 

 lete, ' Cosmos ' yet remains an unrivalled production. We 

 are not, however, to associate exclusively with this work the 

 grand generalisations and unity of science constituting its 

 main features ; these ideas would undoubtedly have found ex- 

 pression had ' Cosmos ' never been written, and had, in fact, 

 been broached by Humboldt long before even the first volumes 

 had appeared. The book stands out unquestionably as the most 

 comprehensive compendium of modern science, and as the most 

 complete history of its development that has ever been at- 



1 ' Briefe an Varnhagen,' No. 17. See ' Kosmos/ vol. v. pp. 131-4. 



2 * Briefe an Bunsen/ p. 139. 



3 Buschmann in ' Kosmos/ vol. v. p. 99, &c. 



