®ite 
gtitanical ^ alette* 
EDITORS: 
JOHN M. COULTER, 
WABASH COLLEGE, CRAWFORDSVILLE, INO. 
CHARLES R. BARNES, 
UNIVERSITY OF WISCONSIN, MADISON, WIS. 
J. C. ARTHUR, 
PURDUE UNIVERSITY, LAFAYETTE, IND, 
December 5.1890. 
Iv^ dear Dr. Watson:-- 
Do not be alarmed when you see the enclosure in this let- 
ter.. I am not going to ask you to revise every page of my MS, but I 
v/ill trouble you to look over a little more. The pacl'.age of letters 
as Y/ell as the three catalogues came duly to hand (and I suppose the 
copy of Henderson will be in on the mail that sees this go out, as I 
have your p. c. announcing its coming.) I have considered those letters 
and am of opinion that ttere are some suggestions that are worth car- 
rying out, IF THEY DO NOf IMVOLVE TOO GREAT A DEPARTURE PROM THE ORIG- 
INAIj plait OP’ THE BOOK. Some oi^them. coincide closely with my own vievfs 
as to desirable changes. In working out the Ranunculaceae ,ho’wever ,I 
took special pains to make as few changes as I tkuDcgkat kkuM could. 
After pondering over the matter for some time I have made out the or- 
der Magnoliaceae after a somewhat different plan, and one that embod- 
ies most of the changes that I think desirable. Those changes are as 
yo\i v/ill see on examining the slips: l.The making out of a key to spsx 
sisK genera to follov/ the description of the order, the choices of the 
key to hiave but one or at most but two characters. V/here¥er possible 
this key is to follow or rather to shov/ the order in v/hich the genera 
will be arranged. I exrect that this cannot alv/ays be done, in which 
case the key would be more or less artificial. 2. Giving a fairly , full 
description of the genus as in the Manual,!, e. under each generic head- 
