942 



SCIENCE 



[N. S. Vol. XXXI. No. 



whole breadth of the page or only half- 

 breadth ; and to do this it is worth while to 

 sketch the figure roughly on the scale that 

 it will have in the text. When this is 

 settled, the figure should be redrawn on 

 double scale with really black ink in smooth 

 firm lines, so that it may be effectively 

 reduced in making a black and white 

 "process" cut. If any lettering is in- 

 cluded, let it be plain and unshaded. 

 The number and title of the figure ought 

 not to be drawn on it or below it; both 

 can be set up in type, when the figure is 

 printed in its proper place in the text, thus 

 saving in time and gaining in appearance. 

 These are trifles: but trifles ought to be 

 properly attended to, and not neglected. 



In addition to the various cartographic 

 and pictorial aids thus far mentioned, let 

 me call special attention to the device 

 known as block diagrams, or bird's-eye 

 views, such as Figs. 2 and 3, which may 

 be designed so as to form useful supple- 

 ments to descriptions that open with con- 

 densed block statements. Both tell the 

 plot of the whole story at the beginning, 

 and thtis allow the reader to place all 

 details where they belong, when they are 

 met in later paragraphs. Just as block 

 diagrams aid in giving graphic illustration 

 to the members of series of deduced type 

 forms, as has already been mentioned, so 

 they aid in the understanding, the descrip- 

 tion of actual regions, because they serve 

 so immediately to present the generalized 

 type forms with which the observer com- 

 pares the actual forms. When seen corner- 

 wise, block diagrams have the advantage 

 of presenting two structural sections, if 

 desired, in immediate association with the 

 surface forms that have been carved on the 

 structural mass. When drawn in groups, 

 they have the further advantage of com- 

 pressing into a single view the several suc- 

 cessive stages of development, which are 



verbally presented or implied in the state- 

 ment of the text. 



Diagrams of this kind are not and are 

 not meant to be mere pictures of observed 

 landscapes, for they must always be simpli- 

 fied by the judicious omission of much un- 

 essential detail, and greatly compressed by 

 the omission of many repetitions of similar 

 elements. They may indeed be rather 

 fanciful, in being designs rather than 

 copies of nature, as is the case with Pigs. 

 2 and 3, above. They should be simply 

 draAvn so as not to demand too much time 

 in preparation, yet they may still be vivid 

 and effective in aiding the reader to grasp 

 the meaning of the writer. 



No one may be more conscious of the 

 defects of diagrams than the one who has 

 drawn them. In the imaginary view of the 

 dissected coastal plain south of Ancona, 

 here given in Fig. 2, the hill shading is 

 very rough; all the slopes are drawn con- 

 vex, and hence fail to show the graceful 

 concave lower sweep down to the valley 

 floors. The terraces in the main valleys 

 and the narrow belt of oldland included in 

 the background are too definite and dis- 

 tinct. The absence of all indications of 

 forests and fields, of villages and roads, 

 gives an impression of barrenness and 

 vacancy that does no justice to the pleasing 

 reality. Moreover, the dissected hills and 

 the broad valleys of two consequent 

 streams extended from the oldland do not 

 coi-respond to any particular hills and val- 

 leys of the district concerned ; they merely 

 show the observer's generalized idea of the 

 kinds of hills and valleys that characterize 

 the district. Nevertheless, the drawing 

 has a value in immediately presenting the 

 essential features of a late maturely dis- 

 sected plain, in which the streams and val- 

 leys are prevailing consequent, with some 

 insequent branches ; in which the hill sides 

 are all reduced to gently graded slopes; 



