600 BOOK EECORDS. 



approximately all, — that has been ascertained respecting the 

 areal and more local distribution of each and every plant, he 

 must needs accept on trust the greater portion of what has to be 

 included; that is, he must learn it through the eyes of other 

 botanists, accepting on faith what they may report to him indi- 

 vidually, or may have put into the form of printed records for 

 general use. 



During the present century, and increasingly during the last 

 thirty or forty years, the facilities for acquiring a knowledge of 

 individual plants as species, and the facilities for visiting their 

 actual localities, both have been vastly augmented. The general 

 result is, not only very numerous additions to our stores of 

 knowledge, but also an increased accuracy in the records of its 

 details ; the per-centage of error becomes less and less, as we 

 get nearer to our own times, and nearer to the actual present 

 time. The farther we go back, and the more we depend solely 

 on old book records, or old labels and specimens, the larger will 

 be the per-centage of error. To a great extent, that which was 

 true and accurate a century or two ago will have been subse- 

 quently verified or confirmed, afresh ; leaving a balance of old 

 records unconfirmed, which are gradually coming to consist 

 chiefly of errors. The state of the case may be summarized 

 thus : — 



1. Old records, more recently confirmed afresh. 



2. Old records, not more recently confirmed : 



A. Accurate at their date. 



B. Inaccurate at their date. 



Suppose that we allow only five per cent, of inaccuracies to the 

 old records of the pre-Linnean times. This would be a small 

 per-centage, because we must let that per-centage cover errors 

 of place as well as of plant, and also subsequent errors in trans- 

 lating the old name-descriptions into Linnean binomials. Take 

 a thousand old records, which will thus give us fifty inaccuracies 

 of one kind or other. Say, that nine hundred of the old records 

 have been subsequently confirmed afresh. There will remain 



