56' ILLUSTRATIONS OF INDIAN BOTANV. 



T 



These little resinous deposifs occupy a prominent place in DeCandolle's characters both of 

 fril)es and genera. Tausch, who strongly objects to l>C'iindolle's (jeneral arrant,-f>nient, especi- 

 ally hi.s suborders, and proposes a new one for the arceptan<-e of liotanists, excludes them froi-i 

 his sectional or Tril)al characters, but adopts them in hi.s trfneric ones Ne <b>rives the char 

 aotprs of Ins tribes from the external forms of the seed, alroirether rf^jectinr- DeCandolle's sub" 

 orders requnmg dissection for th.>ir determinHtioM,and as beir.tr jnron.slant and not always aonli- 

 cable to the species referred to them. That his distribution i.s i, ally an improvement on the ar- 

 rangement he wishes to set aside, I am unable to say, mv collection' bfincr too small to admit of 

 my entering into a comparison, but I suspect both will be found defective when (he order is 

 better understood. One point in both, which I consider objectionable, is the great rumber of 

 SPctioMs and subsections -3 suborders andserentpon tribes inDeCandolle's and-I2 tribes and 

 ili2 «« t^f!" T " ""r^Tf "!• ^^^'' distinctions between these, are often so exceedingly 

 t!dl o \1}^VT ^"^PP^«!^^"r''^' \ ""^^ f''' ^-'''' "^'^ ^^P«^'^'"y d*^^-^««i themselves to tli 

 o nv^s ilation w Ih-^'' '" '1'"- ""^^ l"-^"" ^'"^ ^'''''''' characters. The circular method 

 cipntlv ex^i^^^^^^^^ now rendering such important services to zoology, has not yet been suffi- 



to Lollies as the tr^'; T^^ "^"'^ T"^",^^' ^"^ '^^ ^'"^'dation of ^ich extensive and na- 



a anle^^^^^^^^^ f'> produce a satisfactory 



?he rarr-simDh ^ C^^^^e^^m.e, whether artificial or natural have fallen short of 



dete m .e th^in rilifvaTuVy;^' T" '^' '"f^"'' ^^^^'"-^^ "^^'^'^^^'^ '" ^^^ fi'-^t instance, to 



subdiv ils and in th^ , '^'T^'"^'' ^^^^J' «™P'-^^d i'^ the formation of their sectional 



u ces iTallsyst m bu^^^^^^^^^^^ '^- ''"" ^'"""V '^« this knowledge is indespensible to 



to succeed wUif^^^l^f rnt^ThlJpSli^feS? '' ''' '''^- -' ^ '^''^ 



even hirm^s^^dX^n^f^^f LZrrr'afb "^'^'"1" ''' ^^T^^^' "'^^^^ ^^^ ^-" ^^U-^^^ ^^'^^ 



all generic characte ^f^ m the flo^W and f,^,f, '^''''''''' ^''^'l '''^ ?-^ P''-'P'-^ '^ d--''"=^ 

 natural assemblages of Jneralf?!!" u , ' ^^^'''S notwithstanding, to bring together as 



because a unifolrvatT s i'ed to tl" s'' •' T?'' ^'"^ ^' ^'"'^^ ^"^ DeC^idolfe, simply 

 yet to find some one re'u nin>to fhii H ' i "'^ «^lmracters : and I shouUl not be surprised 

 bution. I confess I ^^Z^ll^: ^T^'!:i^::i^^ '}}^ pnmary ^characters of a new distri- 



Dution. I confess I rannnf cnr^r.«n / "*6'^" '*'* iiiepnniary characlers ot a new clistri 



rangemenf, ,hor;h h „k ve vT f,.r """'TP' will succeed i„ pro.Iurinj . „a.ural ar 

 Thepropo al of^Tau ch tha Vf lu ''°'''''';'''^" ^'"'"'Xi" ™»y l>« obtained f.om (he involucra. 

 fruit, I think good, hut ,; far a9 Itl T '"'f f^'^V"" ''""^ "" """""1 r"""' <'f "■' 

 Burnet gives I sketch o "a sMar nil tl .' ^"'''%^';'V' ''' '^''""^'"rs.is carried too far. Mr. 

 apprehended. That the, are h" htt '. 1 ' 'l iT'^k'*" '"'i"''''^ characters bei.,K n,ore easily 

 >nine, but as wire drawn LLfii;h„I^^^^^^^^ ^r"" ' «"' "™''''^ ""l''"; 



--, . ucmamg a numher of useless mrli,.,',l»r. <•"""■ '^""aciers, or rallier geu.r c i wcripli- 



but which when accidentally wan in^ „, CeH '"", f""i"P» '» «"'J «P'"i^>' "f '^' "^'''"•■ 



spec.es to be distributed as new "ped;s in iXrl " '''^"", °^ '=^"^"'g '«"-•!«=■- "f '!'« »»"'= 



tmc genera. That errors simil»?f„t1.,,r ,^'5"?' 8^".™. »"d "en raised to the r.nk of dis- 



