73 



portunity for any really important details of technique; and 

 charts (or, preferably, typewritten sheets for distribution) offer 

 more economical and more serviceable methods of indicating 

 such details. 



(3) Those on the program often ignored the real uses of charts 

 and diagrams: (a) as visual aids, (b) as time savers. Long lists 

 of names, substances, etc., given orally, dull the attention of even 

 the most interested. But charts should be allowed to speak for 

 themselves. No speaker has any right to hold the audience while 

 he reads every column or describes every curve. He should pay 

 his auditors the compliment of recognizing that they can read, 

 and should not persist in droning over a chart minutes after they 

 have exhausted its possibilities. Years ago in this same Phila- 

 delphia town, a friend watched two children at play in a back 

 yard. In her usual slow way, the older, a little girl of seven, 

 began to tell about a kitten. Her playmate, a boy of six, fidgeted 

 nervously, anticipated every word as she drawled out," Willie^ 

 once-I-had-a-lit-tle-kit-ty,-and-once-it — •" Here Willie jumped up 

 nervously, almost shouting, "Did it bite? Did it scratch? Did 

 it run away?" How I would have welcomed a Willie during 

 some of the papers! 



(4) Contributors too often insisted upon laying before us their 

 day-book instead of the ledger; indeed, the balance sheet itself 

 would often be preferable. Would it not be better to write our 

 papers as we read the articles of others? Usually we turn first 

 to the summary and conclusion, glance back to make sure we did 

 not mistake Alfred J. Smith for Alfred M., who really doesn't 

 know at all anything about the haploid chromosomes (or conical 

 horns, or the entropy of vaporization), and then search in the 

 appropriate parts of the paper itself to see if the striking differ- 

 ences noted are supported by a sufficient number of instances or 

 experiments, or if this result is based on Smith's former method, 

 which, you pointed out last year, was faulty in not recognizing 

 such and such relationships; or else you measure his conclusions 

 by that recent brilliant discovery of Brown's which promises to 

 remodel all our theories and most of our methods in such research. 

 Even a murder trial — or a case of petty larceny — is conducted in 



