92 VARIATION IN LEPTINOTARSA. 



is the form rubicunda, shown in fig. 3, plate 16. This I have found in nature 

 and reared in experiment. It breeds true to type. 



An interesting extreme variation is that represented in fig. 2, plate 16, 

 described by Jacoby as L. angustovittata. I have found this form in nature 

 and reared it in confinement from L. undecimlineata (plate 16, fig. 1), so 

 there can be no doubt of its character. It breeds true to type. 



L. decemlineata has proven a species from which several of these variations 

 have arisen, some of which I have shown on plate 16. Six undoubted cases 

 have been observed to arise from this species L. pallida (plate 16, fig. 7), 

 defectopitnctata (plate 16, fig. 6), minuta, tortuosa (plate 16, fig. 9), melan- 

 icum, and rubrivittata. These, excepting tortuosa, have arisen several times 

 in nature and in experiment and are known to breed true to type. 



It is highly probable that additional variations of this class will be found as 

 observations are extended, and I suspect that many of the species of this 

 genus, which are known only by one or two specimens from one or two 

 restricted localities, are in reality variations of this character and not perma- 

 nent species. 



THE LAWS OF VARIATION IN THE GENUS LEPTINOTARSA. 



From the study of variation in this genus certain rules or laws have been 

 found, which are followed most rigorously by the different species and groups 

 of species. The study of variation can not give us data as to why a given 

 species obeys certain laws, or how modifications are produced therein ; it can 

 furnish only exact data as to their existence and extent of influence. It is 

 important that such information should be obtained concerning a form if 

 further investigation of the problems of evolution is contemplated. 



(1) Direction of variation. In all of the species and characters examined 

 in Leptinotarsa variation is determinate, and in few directions, no case of 

 indeterminate variations being found. 



(2) Certain special laws appear, which apply to this genus and perhaps to 

 others as well: 



(a) Coloration shows variation in two general directions only, toward the 

 dark and the light, the dark being melanic or metallic and the light albinic, 

 xanthic, or rune. In any lot of variates modification exists in not more than 

 three directions, two of which are light, are closely related, and intergrade, 

 while the other, the dark, is diametrically opposite. 



(b) In the elements of the color pattern there is a tendency for the spots 

 to spread out or contract peripherally and the stripes and bands to extend or 

 contract at their ends. The spots, stripes, and bands are most variable in the 

 posterior or distal portions of the structures on which they occur, and least 

 variable in the anterior and proximal portions thereof. Increase of pigmenta- 



