TRKFAOK. 



irmiU\st dirticiilty in (l(>cidiii.u- (lucstions of this kind is in the case of 

 hisidar forms, anions which occur every possible degree of difference 

 between rchited forms inhabitint^- different islands, so that it not only 

 becomes hirgely a matter of individual judgment as to which should 

 be oiven specific and which subspecific rank, but furthermore the dis- 

 tinction made nmst, in the case of any author, necessarily be more or 

 less iii'bitrary, since no ''hard and fast ruhv' for determining- such 

 <|ucsti()ns seems possible. 



As ot)served l)efore, the more familiar one becomes with thesul)iect 

 through the medium of specimens representing continuous geographic 

 sequence of localities the fewer in number really distinct species become, 

 and what have long been considered such resolve themselves, one by 

 one, into a connected series of subspecific forms, each representing a 

 definite geographic area of more or less marked peculiarities of topog- 

 i-iijihy. climate, or other ])hysical features. Such forms are fixed, or 

 '"ti-ue," over territory of uniform physical character, the intergrades 

 coming from the meeting ground of two such areas. Such a group of 

 conspecilic forms may aptly be compared to the colors of the solar 

 spectrum, which form a graded scale from red, through orange, yellow, 

 grcMMi. and bhie, to violet, with intermediate hues of greater or less 

 numl)er, according to the nature of the case requiring their indication 

 by name. These colors of the spectrum, though imperceptibly run- 

 ning into one another, are obviously distinct, and the necessity of recog- 

 nizing them by name has never been-questioned. 



To carry the comparison still further, a certain species may include 

 six sHbspecies or conspecilic forms, which for convenience may be 

 designated l)y the subspecific names ruhra,, avrantiaca^fiava^ rvrldis^ 

 ci/dVKi^ and rlolacea. Intermediates between these might be designated 

 tis riihro-aurantiaca (or cmrmitiaco-rnhr<(^ according to which form the 

 intermc^diate most resembles), aMrantiac<>-flava (or flwvo-mu'aritkuui)^ 

 faro-r/'r/'din^ virldl-cyanea, and cymiw-'Wiolacea.,- i. e., red-orange (or 

 orange-red), orange-yellow (or yellow-orange), yellow-green, green- 

 l)lue, and blue-violet of the color scale. The necessit}^ for such a 

 nomenclature is just as great in zoology (and botany) as in chromato- 

 graphy; and to ignore this necessity is neither scientific nor sensi})le, 

 Init on the other hand is nothing less than suppression or perversion 

 of an oV)\ious truth. The onl}' (juestion that can possibly exist iii the 

 mind of those who have this matter to deal with is the degree of dif- 

 ference which shoidd l)e recognized in nomenclature, and in this respect 

 there is excuse for more or less difference of opinion, accoi'ding to one's 

 ability to discern differenc(>s and estimate the degree of theii- con- 

 stancy, the extent and character of material studied, and the amount of 

 time which has been devoted to its investigation. No doubt many of 

 the forms which the author has recognized as subspecies in the present 

 work may appear trivial to others, especially those who have not had 



