178 EXPERIMENT STATION RECORD. 



tlie readability of its bulletins, marred by this defect. In detailing the 

 results of tests of separators terms are used in this bulletin which 

 may be very familiar to the mechanical engineer, but which are unintel- 

 ligible to laymen, especially when referred to solely by abbreviations. 

 Such a bulletin should certainly be accompanied by a key or an inter- 

 preter. The work reported in this bulletin is of great merit, but in 

 their present form the details are lost to the average reader. 



The character of the tables and their relation to the text are matters 

 which are often too slightly considered by the writers of station bulle- 

 tins. Many readers have an aversion to tables, which is largely to be 

 credited to the blind, involved manner in which data are not iufre 

 quently tabulated. Tabulating requires not a little skill and study. 

 The table should have a title exijlaining what it is about; the cokimn 

 headings should be as brief and plain as possible; as a rule onl}" like 

 measures should be put in the same column, and the table should not 

 cover too many j)oints. It should be so constructed that it will be self- 

 explanatory to a large degree, and will enable the reader to take in the 

 general plan at least at a glance. It is a kind of word picture, and 

 the picture should be so well drawn that the reader will comprehend it 

 without protracted study. Moreover, he should, if possible, be attracted 

 to examine it, and not led to skip it. In a popular bulletin if a table can 

 not be so constructed as to attract the reader, it is at least worth con- 

 sidering whether it should not be omitted or relegated to an appendix. 



With regard to illustrations, we hold that they should not be used 

 unless they serve to explain or amplify the text or to instruct the reader 

 at least as well as words can do. To make a picture book of a station 

 bulletin is hardly legitimate, though if the pictures are first class in 

 conception and execution we can condone such a fault. But to print 

 poor i^ictures which do not really illustrate the text is unpardonable. 

 It is a good rule to make the text as good as you can, and then make 

 the pictures of higher grade than the text. 



And when great pains have been taken to write a good bulletin and 

 to procure fine illustrations, why spoil it by sending it to a poor i^rinter 

 or using cheap paper or bad ink? We appreciate the difficulties under 

 which some stations labor in this matter. It certainly is too bad that 

 local laws or regulations should make it necessary to have the people's 

 literature ruined by bad typography. There may of course easily be 

 extravagance in ijrintiug, but it is only fair that good type and clean 

 paper should be used in giving to the world what it has cost so much 

 patient efibrt to obtain. 



Least excusable is careless proof reading. We have good reason to 

 know how difficult it is to secure absolute accuracy of detail on the 

 jirinted page, but when, as in a recent case, a score of errors in spelling 

 IS found in a list of plants, not to mention the improper use of capitals 

 and italicSj something is wrong. 



