U INTI!( mrCTION. 



lialf as nev.-. Iu so doiug, I sliall be accnsecl of liaviug describecl too mauv 

 iiew species, aiid sliall be censnred for haviug paid biit little atteution to kuown 

 species. But sucb is uot at all tbe case. I liave, as far as necessaiy, cousiiltecT 

 all accessible literature bearing ou my ■work, aud Imve made exhaustive 

 comparisous ■with all the herbarium specimens preserved iu oiu- institute.s 

 at Taihoku and Tokyo. It should be imderstcxjd, as I stated in my former 

 I^aper,* that our herbaria are as yet verj inadequate for the working up of 

 mj materials ■which are reaUy extensive. It frequently occurs with us that 

 foreign herbarium specimens foi- comparison beiug veiy scanty, we are 

 compelled to consult original or reproduced descriptions, instead of herbariuin 

 specimens themselves. 



lu this way have my own investigations beeu made. I have comraenced 

 my work with tlie utmost care, examining all descriptions of the species with 

 which my plauts might be identified. lu those cases in which the descrip- 

 tions of plants uuder examiuation were eshaustive or at least avaiLible for 

 identification, the work could be carried ou witli some measiu'e of satisfaction. 

 Mucli more satisfactory has it been 'when they were aceompanied witli figm^es. 

 When, however, the descriptions were very short and imperfect aud uot 

 accompanied witli any figures, the work of identification was iitterly dis- 

 appointing if not impossible. In cases where I coukl not satisfactorily 

 identify my plauts with auy kuown species in consequence of the imperfec- 

 tion of the given descriptions or the waut of herbarinm specimens, the only 

 work that was possible for me was to describe my plants f ully and acciurate- 

 ly as new species, giving their affinities. This has been most frequently 

 the case with ferns. As stated by Sir Willi.um Hooker,** the identification 

 of fems with given descriptions is a'veiy clifficnlt task, and in some cases 

 utterly impossible, if the descriptions are not accompanied with figui-es. 



For us to whom comparison with herbarium specimens can be but 

 imperfectlj^ carried out, the most important work to be undertaken is not so 

 much the identification of plants with kncnvn species, as the study of their 



* Hatata, B. — Materials for .1 Flora of Formosa, iu Journ. CoU. Sci. Imp. Univ, Tokyo, 

 XXX-1, p. 2. 



** HooKEi;, Sir 'Whj.iam JAtKsox. — A second Cenhiry of Ferns (18G1), p. VII. 



