440 MEMOIRS OF THE CARNEGIE MUSEUM 



toward the head-waters of these rivers and to closely aiiproaeh the divide toward 

 Lake Erie.'^ 



This would favor a direct crossing of the divide h}' actual migration over land, 

 and indeed the river-species are al)le to survive when out of the water for a consid- 

 erable time under certain circumstances, as 1 have ascertained by experiments. 

 During hot and dry weather it is hardly possible to keep them alive for more than 

 an hour or two ; but in cool, cloudy, and damp weathei- I have found that speci- 

 mens suspended on a string on an open veranda"" were not dead after seven hours, 

 and restored to water, recovered entirely. This might at least render a 

 migration over land possible, but I do not think that it actually takes place, 

 since it has never been observed, either by others or by myself, that C uh- 

 scurus, or any other species classed ecologically with the river-species, leaves 

 the water voluntarily. On the other hand it is possible that C. ohscwrus may 

 undergo a passive transport from one drainage to the other, as for instance by Ijirds. 

 However, I do not believe that the crossing of the divide toward Lake Erie is due 

 to the latter cause. It seems to me highly improbalile, not that birds should be 

 able to carry crawfishes for a long distance, but that it should happen that a bird 

 should take up a crawfish in one stream, carrying it to another safe and sound, and 

 arop it there without hurting it. Birds do take crawfishes "^ and sometimes carry 

 them short distances, but this always results in serious injury, even if the specimen 

 is not immediately eaten. Thus, even though we may admit that crawfishes might 

 be transported by birds without being injured, such cases must necessarily be 

 extremely rare, and do not happen often enough to efiect the establishment of a 

 species in a drainage system from which it was originally absent. 



There are other considerations which make the assumption of passive transfer 

 improbable in our case. Toward the east ('. nbscurus is (with exceptions to be dis- 

 cussed below) rigidly restricted to the Ohio drainage, and nowhere crosses into that 



^* At Linesville, Crawford County, I found tliis species in the very headwaters, almost in the springs running into 

 Shenango River just south of Summit, wliich is on the divide. 



*'' Particulars of one of the experiments ( I liave made a series) are as follows : November 9, 1905. Cloudy day. 

 Mean temperature : 31° F. Light breeze from West-South- West, and light snow in afternoon. Specimens of C ohscunis 

 suspended on strings on veranda with southern exposure. Beginning of experiment 9 a. ni. One specimen taken in at 

 2 p. ni., another taken in at 4 p. m , and put into water. Both alive and vigorous ne.\t morning, and were kept alive 

 till December 18, when they were thrown into alcohol. 



In midsummer, on hot days, I often observed that the vitality of C. obscunts becomes very low after they are only a 

 short time out of water. They may die within an hour, without having been subject to any other injury than that 

 caused by the removal from the water. 



5'' Mr. W. E C. Todd informs me that remnants of crawfish are quite usual in the nest of the kingfisher. 1 have 

 seen, in the collection of the Department of Agriculture, Harrisburg, a specimen of C. barloni, taken from the stomach of 

 a kingfisher. 



