164 PRINCIPLES AND CONCLUSIONS. 



Macroglossa and false flowers. — In order to throw light upon the 

 numerous observations of visits by Macroglossa stellatarum and other in- 

 sects to artificial and painted flowers, Plateau (1906:141) collected all 

 the known cases of this sort, commented on them critically, and carried 

 out several series of experiments to show the slight value pertaining to the 

 observations. The significance of the latter and of Plateau's objections 

 is discussed later (p. 189), the experiments alone being taken up at this 

 point. The flight of Macroglossa is astonishingly swift, Bedel noting an 

 average of 50 flowers of jasmine visited per minute, and Plateau observ- 

 ing 301 flowers of Phlox paniculata sought in succession. In the experi- 

 ments with colored cloths, rectangles of silk, wool, etc., varying in size 

 from a 12mo. to a 4to. page and ranging from red through yellow to blue, 

 and rose, were attached to four stakes placed at the corners of a large 

 bed of Delphinium ajacis. The 7 experiments with varying combinations 

 gave the following results: (1) the 12 hawk-moths observed always flew 

 directly to the flowers, never to the cloths; (2) no moth paid any real 

 attention to the cloths, two only making a rapid exploratory curve about 

 the blue woollen fringe; (3) none of the other insects noticed the colored 

 cloths, not even the blue fringe. In the next series, the objection raised 

 earlier by Perez was met by using small pieces of cloth 4 cm. square, which 

 were hung near the flowers of Dianthus and Phlox. In no case did the hawk- 

 moths go to the bits of cloth, even bounding over them as so many ob- 

 stacles to visiting the flower. Similar results were obtained by employing 

 pieces of colored paper, 14 by 10 cm., a single Macroglossa, and one of Papilio 

 and Bombus merely inspecting the papers. When screens covered with 

 paper on which bright flowers were painted were employed, no hawk- 

 moth gave them the slightest attention, and this was true of a large number 

 of bees and butterflies. Two butterflies, two bees, and a wasp merely 

 alighted on the screen for a moment's repose, as shown by the fact that 

 one of them landed on the unpainted side, while one Apis and one Vespa 

 flew up and down along the screen. Four experiments were made with a 

 large number of artificial flowers, which permitted the study of the behavior 

 of 6 individuals of Macroglossa. At least 100 such flowers were placed in 

 two groups near a mass of Anchusa, and at a distance of a meter, and in 

 the second experiment they were scattered in and about a border of 

 Dianthus. While 99 flowers of Anchusa and 110 of Dianthus were visited 

 by hawk-moths, no notice was taken of the imitations, and a similar result 

 was obtained in the final experiment. 



Plateau refrained from giving formal conclusions at the end of this 

 paper, partly because his studies, though prolonged and laborious, were 

 rendered incomplete by the destruction of the gardens in the midst of 

 which he worked. 



"New researches, by means of other methods, will be necessary and I hope there 

 will somewhere be found a serious investigator to undertake them, rather than a super- 

 ficial observer content with approximations. I have conscientiously endeavored to 

 deceive Macroglossa by means of colored paper, large and small pieces of cloth, flowers 

 painted upon wall-paper, artificial flowers, and the colored bracts of Salvia horminum. 



