184 IOWA ACADEMY OF SCIENCES. 



While the work is proceeding the pen becomes empty. 

 The operator must have some means of watching the pen 

 and the paper just below it. For this purpose a mirror, 

 E, is attached at an angle of about forty-five degrees to 

 the plane of the drawing-board. This reflects the image 

 of the pen toward the operator's telescope. Occasionally, 

 the pen must be cleaned. This can be done without dis- 

 turbing its position. The width of the line can also be 

 regulated. The hinge is made to fit closely in order to 

 prevent a motion of the pen from side to side. To provide 

 an additional safeguard against this possibility, a strong 

 wire spring, not shown in the drawing, is attached to F so 

 as to keep the end of the hinge drawn always to one side. 

 The pen can be moved up or down by gentle pressure but 

 not from side to side. It follows any slight elevation or 

 •depression in the surface of the drawing-board. 



One of the most perplexing of the problems that had to 

 be solved was that concerning the drawing-board. Sup- 

 pose we had a board seven feet square. In a damp atmos- 

 phere it would swell more in one direction than in the 

 other, which is a very undesirable property in this case. 

 Suppose you made your board in two sections, an upper 

 and a lower, having the long fiber at right angles. You 

 would be adopting the method of the man who made the 

 wooden saddle. The surface of the board would not be 

 a plane. After some deliberation the following method 

 was adopted: several pieces of broad pine flooring were 

 taken and a line drawn down the middle of the broad 

 side. Fifteen of these pieces were laid side by side so as 

 to leave about three millimeters between each board and 

 the adjoining one. Over these was placed a second layer 

 of boards with their medial lines also marked. Wherever 

 two of these medial lines intersected a screw was driven. 

 In all about 230 screws. This gives a board which will 

 hold its shape sufficiently for the present purpose. The 

 expansion in the direction of the long fiber is very small. 

 At right angles to this direction let the wood shrink or 

 swell for we have given it sufficient room to do so in the 



