SI 4 VARIABLENESS 



some have 10, others 5, stamens; some have 5, 

 others 3 or 2, styles; in the Ricornes of Linn. 18 

 (the Rhododendra 50, and Erica 5 !, of Jussieu) the 

 differences between 4 and 5, 8 and 10, or 5 and 10, 

 are so frequent, as to cause great trouble in classing 

 these plants, after the Linnsean artificial system. But 

 the instance of an inferior germen in Vaccintitm, is a 

 wide and remarkable difference, of extremely rare 

 occurrence, between that genus and it's near relations 

 Menziesia, Erica, Arbutus, &c. I 



In general, variations or diversities of structure have 

 been thought to take place most in the parts of the flow- 

 er, and especially in those accessory, rather than essen- 

 tial, organs, the calyx and corolla. The production of 

 the fruit and seed, especially of the latter, being the 

 main object of all the rest, many botanists have, rea- 

 sonably enough, concluded, that the peculiar organiza- 

 tion, and even the number of parts, in the seed-vessel, 

 and, above all, the form and number of the seeds, were 

 likely to furnish indications of the most important and 



invariable principles of affinity or distinction. Even 

 the diversities in the internal parts, or materials, of 



a seed, have of late been laid under contribution, for 

 the purposes of methodical arrangement ; as appears 

 from the foregoing explanations of different botanical 

 systems. 



As far as regards the comparative number of seeds, 

 the slightest observation will teach any person, that 

 Nature has not always made this circumstance of 



