228 THE CANADIAN ENTOMOLOGIST. 



butterflies winging their way about over the clearing, I did not see a single 

 mistake made on the part of the sparrows. They had become adept 

 enough in two or three weeks to be able to distinguish a Cicada with an 

 unerringness that was simply surprising, when we come to consider that 

 none of their immediate progenitors could have seen or tasted a Cicada. 



Other bird enemies appear to be very few, and these not over- 

 voracious. Mr. J. J. Harrison, of Painesville, Ohio, saw the crow black- 

 bird feeding upon them in 1846, while the labourers on the Station Farm 

 at Wooster claim to have observed the robin to attack them. A species 

 of Tachina fly seemed to play havoc with the latter portion of the brood, 

 and either owing to this or some other reason, they suddenly disappeared 

 between June 24th and June 28th. On the former date, in the Experiment 

 Station orchard, they were excessively abundant, while on the latter there 

 was not a living Cicada to be found there, while the stench arising from 

 the dead bodies was quite apparent to one walking through the orchard. 



As usual, the injury inflicted was slight, except in cases of very young 

 orchards, and I saw in one case a., to me at least, unique form of attack. 

 This is shown in the plate (fig. 3), and instead of the regular, quite con- 

 spicuous punctures (fig. 2) made by the female for a nidus, she appeared to 

 have simply thrust her ovipositor into the wood, and with no further 

 external wound deposited her ova. 



The distribution of the brood in Ohio is illustrated in the accom- 

 panying map, plate 8, fig. i. 



In its distribution, rivers do not appear to have had much influence, 

 as it will be noticed that in southern Meigs county a small area outlined 

 by a bend in the Ohio River is only partly covered ; in one township, 

 Letart, the Cicada not being found at all ; while a corresponding point of 

 West Virginia comes within the range of distribution, even though lying 

 across the river. From this point the dividing line trends slightly to the 

 south-west, passing north of Gallipolis, and extending to the Scioto 

 River, at a point a few miles above its mouth, but not extending beyond 

 this to the westward. North the line follows the east bank of the river 

 until the bend between Waverly and Chillicothe is reached, when it 

 crosses the river and holds to its nearly northerly course to near Circle- 

 ville. Here the line makes a sharp curve to the north-east to the city of 

 Lancaster, in Fairfield county, but trends north-west to the eastern line 

 of Franklin county, thence almost northward along the east line of 



