344 FRANCIS H. HERRICK 



four hours, inspection showed that virtually only a beginning 

 of the foundation had been made, more material at that time 

 lying on the grass or hanging below the level of the bough than 

 actually resting upon it. Yet it is important to notice that 

 with all the molding and turning movements repeated hundreds 

 of times with the stubble and mud employed, tw^o things had 

 been accomplished, which probably could not have been done 

 better in any other way; an adherent foundation had been 

 laid, and compensation for the natural inequalities of the site 

 had been properly made. Had the nest branch showed a much 

 greater incline (compare the hummingbird's nest, fig. i6. Part 

 II), this compensation would have been all the more marked. 

 It is plainly apparent that the adjustment of the nest to its 

 site is a gradual process, and must mechanicall}^ follow as a 

 result of the molding and turning movements. It calls for no 

 planning or reasoning powers whatever, but begins with the 

 laying of the first straws, and is completed with the finish of 

 the last mold. Another question involved is the disposition 

 of material, which will be considered presently. 



On the second day (June 15) observations began at 8.26 a. m., 

 when but little change was apparent in this nest. During the 

 first hour, eight visits were recorded, in course of w^hich mud 

 was brought seven times from the creek, and stubble from the 

 garden once. At this time a rising nest wall began to be ap- 

 parent from my position, and my attention was then directed 

 to the alternation in the turning movements already referred to. 

 The ten visits which followed, all of which I was able to record 

 without interruption, are given with other pertinent data in 

 Table VI. 



Alternation in Turning. — It will be seen that there was perfect 

 alternation in the direction of turning, even when the time 

 spent away from the nest lasted 24 minutes, a fact which can- 

 not be attributed to chance. The number of molds made at 

 each visit varied from 23 to 53, and the number of complete 

 revolutions described from two to four. The time spent at 

 the nest lasted from 2^- to 11 minutes, and the time away when 

 searching for material or otherwise engaged from i^ to 24 

 minutes . 



As the table indicates, both dead grass and stubble as well 

 as mud were used at this time, but an examination of the nest 



I 



