1922] SEARS—TARAXACUM 439 
fe It will be noted that absolute 
By) 9 % 8 2 | Jeaf area decreases rather uni- 
°o aed o @ - 
ae) “>>| formly in SC, a seedling plant 
oe : "| grown under apparently fa- 
lt... 28% | vorable conditions. On the 
Fl ge Ss + 2 * | other hand, SD, a seedling 
Sz . that started under hard con- 
oe 8 : ditions (tight, dry clay), shows 
yes t : ‘ a gradual uniform increase. 
1 Oe eg Of course these facts are 
5 ee ee established easily by familiar 
oa oe a ie observation, but they are men- 
= ee a tioned here to show the posi- 
:: "j}av 5 & © * | tive necessity of studies based 
4 ° band, Contig. One® 
ei g wy] upon field material growing 
gs 4 “oe 6 ae under varying natural condi- 
fs >i. & :  . 1] tions. Not only here, but in 
Sa oh te ee connection with studies of 
ss < <g>, |~SCXylem capacity, very errone- 
2 ee “| Po : “| ous ideas of senescent correla- 
_ F age a 3 % : tion might have been obtained 
8 ele & 3 2 "| by limiting studies to uni- 
a 3e. | __ formly grown culturematerial, 
5 an.) ae as a glance at the graphs will 
2 ee, a show. 
: . ney . Total and mean xylem 
S 7 |. duct capacity measured and 
6 - computed as stated show a 
S$. es oe oe seeming correlation with de- 
“|S5 % 28 : | gree of senescence in the cases 
Boe. py | «(Of SB and SC seedlings, and 
: 9 23 2, = :.| «somewhat less in the case of 
»i§ 8:2) &: | RB, a secondary rosette. The 
:g gq : :2/88 : | possibility that this correla- 
s3e ps El 2: | tion is real, however, is quite 
Bo _? gta a3 g | upset by data on the depau- 
¥eS%EXS) 4 |  perate seedlings SA and SD. 
