240 Bulletin 133. 



march and at work. In most cases, the caterpillars in each of 

 these armies must have been numbered by the millions ; even an 

 approximate estimate of the number ot worms in a single army 

 would have been impracticable. Oftentimes when an army was 

 marching across a lane or roadway, nearly the entire surface of 

 the ground for several rods would be covered with the crawling 

 mass of worms ; one could not step without crushing several of 

 them. Is it to be wondered at that when such a vast horde of 

 hungry creatures reach a field of young corn, acres of the plants 

 soon look like the one shown in figure 68 ? They soon strip all 

 the leaves from the stalks of oats, rye, and similar plants, and 

 often cut ofi" many of the heads, leaving them uneaten on the 

 ground. In one instance, a barn loomed up before the worms 

 directly in their line of march, but nothing daunted, many of them 

 valiantly scaled the perpendicular wall and soon succeeded in 

 getting over the eaves onto the roof. Here, however, they met 

 their Waterloo, either from the exposure on the heated shingles 

 to the sun's rays or from other causes, and a windrow of dead 

 worms was formed under the eaves. 



The worms feed mostly at night, remaining hidden during the 

 day in the soil or under chunks of dirt or anything that will 

 afford protection ; many of them find shelter down in the cavities 

 at the bases of the unfolding corn leaves. On cloudy days the 

 worms usually feed during the day, and often many of the worms 

 may be found at work on sunshiny days. 



In almost every case it was the July brood of worms only which 

 ravaged crops ; apparently in none of these instances did the May 

 or the September broods of the caterpillars attract attention. 

 However in at least two localities (Cheviot, Columbia Co. and 

 Warwick, Orange Co.), it was only the September brood of 

 worms which were numerous enough to be noticeably destructive; 

 curiously enough the insect had not been seen in these localities 

 before during the year. Thus, while in most localities all of the 

 damage was done in July, and in a few places the worms were 

 not injuriously numerous until September, it is to be noted that 

 we have no evidence that more than one of at least three broods 

 of the insect was noticeably injurious in any locality during i8g6. 

 A probable explanation of this noteworthy fact will be brought 



