EXPLORATIONS IN WE8TEKN CANADA, ETC. 129 



f with miii 



than III li LOCAL VARIATIONS.* 



the iiniii'" 

 I most lii- 



SiiUT iill slnirtuics <lill'«'riii;: Iroiii tli»' nviiaj.^f arc usually ronCiiM'd to a dcliiiite 

 horizon oi imin- or less icsirit KmI icyioii, ail sacli (litlt'iomcs may Itc ((iiisidorcd local 

 variations. Tlic lar^ci /.oo<fcoj.';iaitliical icfjioiis or provinces arc in this extended 

 seUNe IcM-alitics. and the oidcrs, lainilics. or species arc llic local vaiiations peculiar to 

 the region or proviiae. A soincwiiat moie restricted dcliiiitinn would include sucdi 

 pheiioitiena as arc noticed in tlic |iecu!iar nindilicalinns ot the tins ot I'acitic slope 

 fVesij-walcr lish dcscnhcd in the prexious chapter. Sonu' <'iiinini<hr oi' the Colorado 

 basin, for iii>laiice. Iiave tlie anterici' dorsal rays srron;i ami spinous, while al! the 

 Atlantic- slope >pccies have Iheiii weak and riidiinciitary. Another iiistame is the 

 increased nniniicr ot rays in tiie lins ot' racilic slope lishes. Still another instance is 

 oB'ered l»y the I't rcnp.siihv. ('uliniihln has stroll;;' -pnics in both the dor.sal and anal 

 tins, while /Vc,(>/m/«. the .\tlantic slope <,^'niis, iias none. 



For the present purpose I want to restrict the incaiiin;; still further. In studying; 



; the South Ainerican catlishcs, I tbiind tiiat all th*- .Viiia/onian species ot' tiie ;:<-nu.s 



Ltlantic. \IihatHdi(( have <» dorsal rays, whih' several of the southern forms have more. One 



I c()ndit lull peculiar to the La IMata has (!-'.(; another from the San I'^rancisco has 10 'ays. ."More 



nations in 'euiarkable still is the »'asc of I'sciKlDpimrhKhis ziniiiHVD. All the specimens taken in 



he Amazon have <> dorsal rays, whileofa iimaller numl)er taken liirthcr .south several 



lave 7 dorsal rays. 



seutcoiiditi' It is to variaticuis like the last, /. <■., variations within the spcci»'sor closely related 



have mall 11 pedes found in ditVerent hicalities within a restricted re;;ioii, that I want to contine 



sloi»e of Ni'iy present remarks. Variatiiuis within species are a matter of lines and curves, 



jinute nieasiirements, and shades of cohu': all matters ditlicidt to keep in mind, still 



I ain not m 'ore so to represent to others. All naturalist.s are aware of the existence of .sliyht 



of the ofii 'ff®rcu<*es peculiar to different localities, but sueli variations are usually luit va^-uely 



onceived l)y the observer, and still more va,!;uely by any one to whom the observer 



and proli; '*y attempt to ex])laiu them. 



7 A ..Will, ;i The past summei I colle(!ted a large series of sitecimens of LvKrisrus and liuhnrtl- 



mim. These were taken in a number of dilVerent localities and in two separate 



ver systi'Mis. the ('olumbia and tiic I'raser. The localities extend from tide water 



t an elevation of 1,1I(I0 feet on tlie Kraser, and from 'MM to l.TOt) t'cct on the Columbia 



.'Stem. 1 have also examined a niimber of si>ccimens collei fed by l>r. .Ionian in 



tab. There weie in all H'.u; specimens which I was personally ahlc to examine. In 



lese speciiiu'iis the local vaiiations are so well marked that a ;;rapliic method of 



Mnonstratin^i' the variations i.s possibh'. 



Before attempting t(» explain the charts which illustrate this matter, it is ncces- 

 ry to state that tliere iiave been known from tli two river systems two {groups or 

 ■n era of ('i/ih-inidir haviiif;' elongate anal tins. lesc were h'i<liiiril,si>tiiiis [hiiUcdtKs 

 d lateralis] and a section of Lend. sens iiiioiitxinis, hi/droiihld.r, and iiilli). There are, 

 st, ■variations which do away with the genus I\'ich<tr<lN(iiiiii.s, as distinct from /Aitcis- 

 s; second, a number of variations wiiich, while very striking, need not be taken 

 o consideration, because tlie \'ariations in a single character are siitlicient for our 

 rposes. We shall limit the ob.servation to the variation in the number of anal rays. 



' liCiid at the Doofiulicr iiifotiiit;' of llic Imliaiiu .VcaiU'iiiy ol'ScioiiLe.s, 181(2. 



F. e. 1'.. ls'Jl-0 



