70 THE TEACHING BOTANIST 



even in certain phases of general natural history study; 

 but they reflect not at all the clearly cut ideas which 

 should characterize the activities of the laboratory. 

 Samples of clear diagrammatic drawings may be seen 

 later in this work (Figs. 12, 13, 14). Practically, the 

 best way to make these freehand drawings is first 

 to outline them very faintly in pencil, and then alter 

 this outline until it corresponds to that of the object, 

 after which a single, firm, complete, even line can be 

 run over it, and all the lighter lines erased. All 

 mechanical helps, such as rulers, compasses, etc., 

 should be allowed when they contribute to accuracy. 



An important principle in making the drawings to 

 illustrate the structure of an object is economy of 

 number ; as many drawings should be made as are 

 necessary fully to illustrate the object, and no more. 

 Thus, for a seed like the bean (Fig. 12), two draw- 

 ings are sufficient ; an end view would bring out little 

 if anything not already in the other two. One view 

 of an object need not duplicate what is already in 

 another, though different views of the same feature 

 should always be included. The extreme aspects of 

 an object should be chosen for representation, i.e. a 

 face or edge view should be an exact face or edge, 

 and not a quartering view. Of course, different draw- 

 ings of the same object should be perfectly consistent 

 as to size, form, etc. 



The scale of the drawing in comparison with the 



