NOTICES OF BOOKS. 



153 



duced a multiplicity of liereditary races, but the characters of their 

 flowers and fruit are still those of Bramca. So with the races of 



w 



Triticum, these may differ in the number of seeds and their form, and 

 in their awns (beards), but no one has proposed to make a new genus 



of any of tbem.* 



The fact that isolation is one of the conditions that leads most usu- 

 ally to the subdivision of species or formation of subspecies, if taken to- 

 gether with another foot, that the majority of analogous species are aggre- 

 gated within more or less contracted areas, appears to M. de CandoUe 

 to militate against the hypothesis that time and isolation may account 

 for the origin of species. Thus the hundreds of Cape Heaths, he says, 

 cannot have owed their origin to geographical isolation, for we cannot 

 conceive causes that would, after their segregation, have aggregated 

 them again. The genera Slylidmm, Solamim, Aster, Astragalus, Cistus, 

 and Linariaf are quoted as affording parallel cases. 



Upon the whole M. de CandoUe inclines to admit two modes of ori- 

 ginating new specific forms : the one derivative, which is very rare and 

 - confined to species that are very closely allied J but geographically 

 sundered; the other an original creation, "mode par une formation 

 propre," which certainly obtains for the immense majority of species. 



uiis are pursued, the genera ot (Jraases rnusi ue reuuceu tu vcij i<..., <..."" """"" 

 of small genera around Trificum itself must be eliminated. We do not admit bearded 

 aud beardless Wheats to be different genera, because we know their history too weU, 

 and not because they do not in the abstract present good generic characters ; tov 

 there are many genera of Grasses contradistinguished by all botanists by those very 

 cLiracters, and which present species that vary so that they may be referred indis- 



J - - ^ ^ 



criininately to any of them. 

 ^ The ' 



T ine very same class ol tacts is auuuucu, uj =>.." „™— , - — - . 



of many si,ccies out of one, in support of their view ; and in conjunctioa wi h he 

 fact that all these genera, except perhaps S/j/lidua» present heaps of scarcely d^tm- 

 guishable species ts subspecies or' races, are no doubt apparently «ro°g° support 

 of it ; to which may be added, in the case of Cisiu^ and Encji. the facibty of hy- 

 bridization and impLsibility of tracing the parents of many of our garden hybrid . 

 If instead of citing those foreign genera we take some X. Si S/f about 

 jenting groirps of k-P^icaUv ^Jg^^^ :^^'£ et^if ap^;af:s:'trfng?; 

 Str ^^J^ar ^ ^ S" ^ an ^.ent inprod^n^speei^^r 



mibspecics. ETen time does not seem to be ^'^'^'^'t'^XlZ nSr BmnwS S 

 mongers find new species of Willows in modern osier-beds and new Brambles by 



'j Pm b. argued W »», .Hat if .he "^t" oT" ^.h'Sf.nS SLt 



E/rr^:^." „tr r £-.^ r r;ra t^ j^, w^^^ 



tnry^o one another. Isolation is in its operation synonymous with alKrcd^s 



VOL. VIIL 



