648 Original Articles. [Oct., 
But on the other hand it must be borne in mind that the 
differences between the several atomic weights compared with one 
another, are for the most part not exactly but only approximately 
multiples of 4; whilst in a few instances, at any rate, the approx- 
imate difference in atomic weight between closely allied elements, 
is not 4 or some multiple of 4, but 2 or some odd multiple of 2, and 
in other instances even 1 or 0. 
Since many of the elements occupying analogous positions in 
different groups haye closely approximating atomic weights, it is 
evident that the mere determination of the atomic weight of a newly- 
discovered element assists us but very little in deciding to what group 
it belongs, but only indicates its position in the group; since among 
the members of every well-defined group the sequence of properties 
and sequence of atomic weights are strictly parallel to one another. 
Doubtless some of the arithmetical relations exemplified in the 
foregoing tables and remarks are simply accidental; but taken alto- 
gether, they are too numerous and decided not to depend upon some 
hitherto unrecognized general law. 
ON THE BUTTERFLIES OF MADAGASCAR. 
By Ronanp Triwen (Cape Town), Memb. Ent. Soc. Lond. 
In the belief that a brief consideration of the Rhopalocera inhabiting 
Madagascar may in some degree aid in the investigation of the relations 
of the general fauna of the island to that of Africa, which has been so 
ably commenced by Dr. Sclater,* I have drawn up the following notes. 
I must at the outset express my regret that the materials at my disposal 
have been too scanty to admit of the preparation of a paper affording 
a complete view of the subject under discussion; but it is hoped that 
these few remarks may prove of service to those who have access to 
ample means of pursuing the inquiry. I would observe, however, that 
the data and observations here given are, for the greater part, the 
results of some years’ study of the butterflies of Southern Africa. My 
obligations to Dr. Boisduval’s admirable work, the ‘ Faune Entomolo- 
gique de Madagascar, Bourbon et Maurice, + are too manifest to 
require comment. 
The total number of butterflies known to me as inhabiting Mada- 
gascar (exclusive of the Mascarene Islands) is 73. These 73 species 
are comprised in 34 genera, and belong to the following 11 families, 
viz. :$—Papilionide, Pieride, Danaide, Acreide, Nymphalide, Saty- 
ride, Hurytelide, Libytheide, Erycinide, Lyceenide, and Hesperide. 
* «The Mammals of Madagascar,” in ‘Quarterly Journal of Science,’ No. 2, 
April, 1864. 
+ Paris, 1833. 
{ I follow the arrangement of Messrs. Doubleday and Westwood’s ‘ Genera of 
Diurnal Lepidoptera.’ 
