New Zealand Flora , with a Reply to Criticism . 365 
ranging on the average 132 more miles (difference i*i, each o-i representing 
13 miles). 
When one comes to examine these figures a little more closely, one 
notices in the first column that 37 wides (56 percent.) cross Foveaux Strait, 
ranging the entire length of New Zealand, while only 9 range the two main 
islands without crossing the Strait. In the second column only 37 per cent, 
cross the Strait, and in the third only 10 per cent. This is as one would 
expect from species picked in order of commonness. But when one goes 
on to the wides without endemics, one finds that even the first column 
shows only 31 per cent, crossing the Strait, and 37 per cent, held up there. 
It is clear that on the average these wides arrived in New Zealand as late 
as, or rather later than, the second wides in the genera which have endemics, 
and in the same way their later species (column 5), though 19 out of 34 are 
second arrivals, are rather later than the third (and later) arrivals of the first 
lot of wides, those with endemics. This table shows, in a very clear manner, 
that the wides with endemics are on the whole earlier arrivals than those 
without. 
One may examine these tables in another way. The 46 wide genera 
with endemics that occur in classes 1 and 3 have altogether 304 endemics, 
those in classes 3, 4, 5, and 6 have 148, 1 and those in classes 7, 8, 9, and 10 
have 7 only; again age is indicated as being more likely to ‘ involve * 
endemics. The 151 swamped genera have between them only 481 endemics, 
against 431 for those 66 genera which also contain wides. These facts go 
to show that on the whole it is the older wides which are accompanied by 
endemics, not the younger, as might be expected upon the hypothesis of 
swamping. 
Dr. Sinnott’s objection that age and area will not explain the New 
Zealand flora because it is based (cf. my diagram in 19 , p. 443) on a central 
point of arrival of the flora, and it is generally agreed that the flora arrived 
in two or more directions, seems to me due to misunderstanding of my 
tables, which show with great clearness that there were at least two sources. 
I have already been' into this question, but may just call attention to the 
fact that though there is a very clear northern invasion, the southern is so 
much larger that the total figures show practically the same result as the 
southern only—the northern are lost in them. 
I am much indebted to my daughter Margaret for drawing the diagrams 
here reproduced, and to Dr. H. B. Guppy, F.R.S., for criticisms. 
1 Including 77 Veronicas and 22 Myosotises. 
