SCIENTIFIC SUMMARY. 
447 
observations, and are obtained by dividing by 10 the product obtained by 
multiplying the number of observations by 2n — 1 (n being tbe order of the 
eclipse) and by 1, 2, 3, or 4, according as tbe observation was recorded as 
fair, good, very good, or excellent. Tbe mean result is 298,400 kilometres ; 
multiplying tbis result by 1-0003, tbe refractive index of air, 298,500 kilo- 
metres is obtained as tbe velocity of light in a vacuum. Tbis, Cornu 
believes, is accurate to of its value. It is a close approximation to tbe 
result of Foucault, 298,000 kilometres, and also corresponds .very closely to 
tbe value obtained from tbe solar parallax, wbicb bas recently been calcu- 
lated by Leverrier, from observations upon Mars and Venus, to be 8"-86. 
Cornu believes tbat with stations separated from 20 to 30 kilometres, it 
would be possible by tbis method to obtain a value accurate to within a 
thousandth.— See tbe ^^Comptes Kendus,” Ixxvi. p. 338. 
ZOOLOGY AND COMPAKATIVE ANATOMY. 
Kingfishers and Fish. — This subject, wbicb is of some interest to 
Naturalists, bas a note upon it in tbe “ Scientific American ” for Aug. 16. 
The writer, dating from California, says tbat Mr. Darwin, in bis last book, 
states tbat tbe kingfisher always kills tbe fish before swallowing it. Dr. 
Charles C. Abbot states tbat tbe fish is swallowed without killing, and often 
while tbe bird is on tbe wing. So far as my 'observation goes, when a fish 
is large, or about two and a half inches long, it is killed before being swal- 
lowed. I once saw a kingfisher light on a limb close to the surface of tbe 
water in a creek ; and tbe bird, having an eye to tbe business on band, did 
not see me (I was about fifteen feet off*). It presently dived into tbe water, 
and returned to its perch with a fish in its bill, about tbe above stated 
length. Tbe bird then began to beat tbe bead of tbe fish against tbe limb 
on wbicb it was standings after a few beats it would stop to see if the fish 
was dead or not ; tbis was done three times, when tbe bead of tbe fish was 
bleeding, and tbe limb against wbicb tbe bead was beaten was stained with 
blood. The fish was dead, and it was then swallowed. Now tbe above- 
named gentlemen may both to a certain extent be correct. The kingfisher 
may swallow tbe small fish without killing them j in my mind there is no 
doubt tbat they do.” 
A new and large Amphipod bas been captured by tbe Challenger expe- 
dition, and is described in a letter addressed to tbe Eoyal Society. Tbe 
author of the description, Herr Von Willemoes-Subm, says tbat, among 
tbe Amphipods known to us, Phronima is its nearest relation. But there 
are so many points in which tbis genus differs from Phronima, tbat it 
cannot form a member of tbe family ; and be therefore pro- 
poses to establish for it a new family, Thaumopidce, belonging to the tribe 
of Hyperima. The form of tbe head is, totally different from tbat oi Phro- 
nima ; tbe antennae are not situated near tbe mouth, but at its front, and 
tbe enormous faceted eyes occupy its upper surface. Tbe first two pairs of 
thoracic appendages are not, as in Phronima, ambulatory legs, but maxilli- 
