I'liivi rsitti of CalifoDiia ruhlicafioiis in Bofami [Yol. 6 





Narboroujih (o9) 20 



Albemarle (205) 71 — 



Jervis (22) 2 



Dunoau (51) 15 14 



Baniugtou (40) 25 11 



Seymour (47) 14 12 



James (153) 40 37 



Indefatigable (70) .... iM -Ji! 



Charles (207) 64 49 



Hood (5il) 2."> 10 



Gardner (33) 12 S 



Chatham (231) 56 45 



Abiiio-dou (50) 24 10 



Biudloe (42) So 12 



Tower (19) 12 3 



In Table TT. Professor Robinson's figures have been converted \\\\o 

 percentages. The islands have also been listed in approximate geo- 

 graphical instead of alphabetical order. Since any number denoting 

 the species common to two ishnuls can be expressed as a percentage 

 of the total number of species on each isknul, the horizontal rows and 

 vertical colunms of figures are not identieal. As one reads down- 

 ward, one encounters the various percentages of the number of species 

 found on the island heading the column, occurring also on the various 

 other islands. Thus, of Xarborough's fifty-nine forms, forty-two. or 

 71 per cent, recur on Albemarle, one. or 2 per cent, on Jervis, nine, 

 or l-'i per cent, on Duncan, and so on. Reading horizontally for 

 Xarborough. however, one eucountei's the figures twenty, nine, eighteen, 

 and so forth ; which denote that the forty-two Xarborough- Albemarle 

 species constitute 20 per cent of Albemarle's total of 205. the two 

 Xarborough-Jervis 9 per cent of the twenty-two on Jervis, etc. 



But these percentages are also unsatisfactory, since it is obvious 

 that when they are read vertically the high figures are regularly en- 

 countered for large and fiorally well-stocked islands, and that when 

 they are read horizontally the figures rim higher just in proportion 

 as the islands referred to are small or poor in variety of vegetation. 

 This is made clear bv the italic and bold-faced numerals introduced 



