453] COLORS OF TIGER BEETLES— SHELFORD 59 



the marking of C. tranquebarica Herbst as the underlying type "from 

 which all forms observed in our Cicindelas have been derived". He 

 bases this statement on the fact that it is the so-called humeral lunule, 

 middle baud, aud apical lunule which give similarity to the patterns 

 of the genus. He states that modification occurs in any one of four 

 ways : 



A. By progressive spreading of the white. 



B. By gradual thinning or absorption of the white. 



C. By fragmentation of the markings. 



D. By linear supplementary extension of the white. 



These tendencies are all recognizable, all of them occurring in the 

 course of individual and geographic variation of single variable species. 



Walther Horn (190S) in Genera Insectorum discussed the patterns 

 from a somewhat different point of view. He states that in the ideal 

 sense the markings which he recognizes as the humeral, apical, and 

 middle spots are made up of 3 humeral, 4 middle and 3 apical spots 

 as shown in figure 290, plate XV, and 333, plate XVI. Thus he calls 

 the markings which are most characteristic of the group the Marginal 

 Component. He calls the median basal spot of the elytron the Basal 

 Component (Bl) and the marking along the suture or anal border of 

 the elytron the Sutural Component. He recognizes also such patterns 

 as those sho\vu in figures 2-41, 243. 248, as Dis^pcrsion Component. He 

 states that this analysis is for taxonomic purposes only and not based 

 on ontogeny. He recognizes the most important tendencies toward 

 joining of spots, in addition to the general plan outlined in G. Horn's 

 four statements. 



The work of these men is here cited to show the fact that various 

 generalizations have already been made sho^ving that the patterns eon- 

 form to a general plan of spots or bands which have been similarly 

 interpreted, though not exactly the same, by two authors with wide 

 experience in the group. 



For the purposes of illustrating what may be determined in the / 

 group in the way of general tendencies (p. 36) and the patterns of 

 interrttpta, intcrrupta subsp. gahonica, flexuo.ia (PI. XII), tranque- ; 

 iarka. and purpurea. And for a second illustration take the same 

 species substituting .^cutfUari.'i for purpurea. 



First noting interrupta and e/aionka, (Figs. 156, 156 a and 165, 

 and 165 a) one finds that the cross bands clearly recognized in Coleop- 

 tera, especially Chrysomelidae, and Lepidoptera. and which appear in 

 the tiger beetle group especially in the patterns associated with inter- 

 rupta (PI. XII), and which appear in all the species in which ontogeny 

 was studied, are present. In gahonica it appears that through iiidivid- 



