218 THYSANOPTERA AND ORTHOPTERA. 



spot. I took a quick step, and noted as accuratel}' as possible the exact 

 point Avhere two neighboring grasshoppers arose. They were so close 

 together that this could easily be done. The distance between these 

 points was then measured. It was further required that the two mem- 

 bers of each pair should be in a line approximately north and south. 

 If they had been taken in any direction whatever from each other, it 

 is conceivable that the grasshoppers might be arranged in long lines 

 bounding open figures that contained few or none of them, so that the 

 apparent number of individuals would ha^'e been much greater than 

 the actual number. I believe this difficulty to have been obviated by 

 selecting insects situated in a definite direction from each other. An 

 average of twenty measurements gave 12.9 inches as the distance be- 

 tween each two grasshoppers. 



The accuracy of the method was roughl\' tested by trying to count 

 the specimens that arose as I walked slowly through the grass. This 

 could, of course, be only approximately done. The insects arose 5 

 feet on either side of me, and at each step I counted all that flew. 

 Locustidae were omitted when they were recognized, as they usually 

 could be by their flight. M\^ steps averaged 2.5 feet in these tests, 

 and the average of a large number of counts showed that 28 grass- 

 hoppers were flushed at each step. There are several sources of error 

 in this method, some of which tended to increase the number counted, 

 others to diminish it. Among the former is the fact that many of the 

 grasshoppers flew forward and were counted at the next step. The 

 numbers were probably diminished, on the other hand, by failure of 

 many of the insects in the edge of the ten-foot strip supposed to be in- 

 cluded, to fly when I moved. If 12.9 inches was the true distance be- 

 tweenHwo specimens, and they be supposed to be arranged in squares, 

 then 21 or 22 should have been flushed at each step. The fact that 

 28 were flushed at each step shows that my estimate probably does 

 not make the number too high. In this case, again supposing the 

 grasshoppers to be arranged in squares, there were about 37,878 of 

 them per acre. 



The amount of food eaten by a single individual was determined by 

 caging a specimen of Melanoplus hivittatus femoratus and feeding it 

 known amounts of grass. Blades of timothy of approximately equal 

 width (4 mm.) were put in the cage after measuring their length. As 

 they became dry, they were replaced with fresh grass, and the length 

 of the uneaten grass was deducted from that put in. That it had not 

 shrunken in length by drying was shown by measuring other blades 

 before and after drying. Of such grass blades, 50.4 inches were eaten 

 in 24 hours. After returning to laboratory facilities, the same aggre- 

 gate length of similar leaves of timothy was collected and allowed to 



