1916] Kroeher: Floral Bclat ions Anwng Ulc (Jala pagos Islands 215 



Albemarle 29 of 325 



James 22 of 224 



Xarborougb 18 of SO 



Duncan 20 of 103 



Jervis 14 of 42 



Sevmour 19 of 52 



Charles 33 of 319 



Chatham 36 of 306 



Hood 29 of 79 



Gardner 23 of 48 



The nearest laud is Iiulefatigable, 14 per cent of whose flora it 

 possesses, as against 12 per cent of Charles', 10 of Chatham's, 10 of 

 James', and 9 of Albemarle's. After all, close proximity counts for 

 more than distinct exposure to the strong southeastern influence. 



GAEDNER 



This island lies close by Hood and fifty miles from Charles; but 

 it h;is forty-four of the larger island's species and only thirty-seven of 

 the nearer ones. 



AYithin its usual limits, however, location makes itself observable: 

 compare forty-four species shared with Charles, thirty-five with Albe- 

 marle ; thirty with Indefatigable, twenty-seven with James ; twenty- 

 four with Duncan, twenty-three with Abingdon — the more remote is- 

 land being in each case also the richer, though less represented on 

 Gardner. 



I cannot explain the low number (thirty-one) of Chatham species 

 on Gardner as compared with the forty-four from Cliarles. Usually 

 Charles and Chatham appear substantially as a unit in their relation- 

 ship with other islands ; and even in the case of Hood — to which 

 Gardner is attached — the difference in favor of Charles is compara- 

 tively slight. 



BINDLOE 



Bindloe is the largest of the northern islands and the nearest to 

 th(^ ceutral group, but, either on account of a lower elevation or for 

 some unknown reason, it has less than half as many speci&s as Abing- 

 don. The affinities of its flora are very evenly distributed, except 

 for somewhat higher percentages for species shared with the other 

 northern islands, as indeed is only natural and might be expected, 

 though Dr. Stewart's cited passage professes the opposite for the 

 northeru islands in general. Thus it has twenty-six of Abingdon's 

 species, as against twenty-seven, twenty-four and twenty-nine of In- 

 defatigable 's. James' and Chatham's, although these average more 

 than twice as many total forms; and similai'ly, the figures for Tower 

 are ten out of twenty-two, as against ten out of forty-two for Jervis 

 aiul el(n-en out of fiftv-two for Sevmour. 



