Page:Condor13(6).djvu/32

200 TH]? CONDOR VoL XIII Imagine the same numbers passing across a line a number of miles long, or better, across the breadth of Siskiyou County, and for say eight hours a day for sev- eral da3?s; the numbers become incredible. The butterflies were all migrating southward. In the early morning none were t6 be seen, but by half past nine they were in full migration. During the night they rested among the leaves on the trees or shrubs, on the sides of buildings or in any other convenient place. With the life cycle taking but a year, it at first seems hard to explain why the numbers should be so much greater one year than another. One factor govern- ing the phenomenon is the presence or absence of fortunate conditions for hiberna- Fig. 70. BUTTERI?LIES GATHERED TO DRINK AT A DAMP PLACE IN THE ROAD. PHO?OGRAPH TAKEN AT SISSON, SISKII/OU COUNT? CALIFORNIA, AUGUST. 23, 1911. tion of the butterfly. As the pupae are parasitized to a large extent, in the neighborhood of 35 percent, and doubtless the larvae are also parasitized, the abundance or scarcity of these parasites must govern the numbers to some extent. A third factor, and without doubt an important one, is the part played by birds in the destruction of larva, pupa, and imago. Many of the pupae, 15 percent or thereabouts, apparently showed the work of birds, and as will be shown birds have an important part to play in the destruction of the butterflies. To what extent birds feed on the larvae is not known. Probably not to as great a degree as on the pupa and imago forms, for the larvae are well protected by stiff hairs. The scarcity or