copyrignt reserve 
r.— . 
bike natural order of Convolvulacece tn u- 
tand from which it cannot be separate, 1 .u h * indl 
W than an embryo in the usual form ’ftl?? the 
i and without cotyledons proper ’v tlp forn 
,ber of genera on subordinate character 2?** ' 
Kipon erroneous observations. ’ as nas * 
^ good grounds for a subdivision of the 
F® the fruit These same characters itm""! ^ 
f in 10 D f U ^ eroua genera, and ev’e n s 
f division of Cuscuta; but even ’ 
his Monograph and in De Candolle's^ T 
^. Splitting Convolves into numerous ge^n 
j si separating the large crowd of species inte T 
mkve no scientific value, and can certainlyTt 
fad, and not too numerous genus, as CusZ * 
are distinct j or they are united in th 
iness throughout, or are thicker at base (sub* 
^of equal length In one group, and ungual■ 
#>d, and of the same thickness as 1 
“ ** are oft en, though wrongly called 
2" • and flatt6ned or U8 °*% impressed at 
** 8ti « matic «“rface of the dilate 
^opening transversely by a regular joi 
J««ged maXT^r i- 
^ther falls out of the persSc^yx, 
