376 



SCIENCE. 



[Vol. II., No. 32. 



so convenient and instructive as the circle, in whicli 

 tlie quantities are measured along the radii, from the 

 centre as their base. The entire radius may some- 

 times represent the whole of that of which these 

 partial measurements are a part. 



For a progressive series, the ordinates representing 

 quantities should be measured from a level base-line, 

 and the scale of graduation shown upon the side 

 marfsin, while the time is measured from left to right 

 by a scale along the upper margin. 



For simple comparison, a series of bars or lines, 

 measured from a common base, and either horizontal 

 or vertical, is a convenient and striking mode of illus- 

 tration, and has now come into very common use. 



A rectangular area, with parallel divisions, amounts 

 to the same thing as a line; but with this difference, 

 that a secondary subdivision may sometimes be repre- 

 sented with great effect. 



Having thus stated some points in reference to 

 graphic illustrations upon a true principle, I should 

 not leave the subject without a word of censure for 

 some that are false. I will specify, particularly, such 

 as attempt to represent comparative quantities by 

 concentric figures, such as circles or squares. The 

 eye has, in these cases, no means of just comparison ; 

 and they are very apt to mislead, where they are 

 intended to instruct. 



The same objection may be made against similar 

 geometrical solids; for, although they may be literally 

 true, their contents being to each other as the cubes 

 of similar lines, the eye does not readily see the 

 difference. It would be better, in such cases, to use 

 cylinders or prisms of the same base, but proportioned 

 in length to the quantities that they represent. 



PAPERS READ BEFORE SECTION I. 



Life-insurance and self-insurance. 



BT EI.IZOK VfHIGHT OF BOSTON, MASS. 



This subject has been a favorite theme with Mr. 

 Wright for several years. His ability as an actuary 

 is acknowledged; and his theory .on this subject has 

 at least the merit of disinterestedness, so far as in- 

 surance-companies are concerned. Without going 

 into the technicalities and mathematical considera- 

 tions that must be met in a thorough review of Mr. 

 Wright's theory, its object may, perhaps, be stated 

 correctly in a few words. A reserve is accumulated 

 by the practical workings of life-insurance in a well- 

 regulated company, which is more than sufficient to 

 meet the claims upon it as fast as tliey mature. The 

 usual system divides that reserve, less the amount 

 which the company withholds as a surplus for ex- 

 traordinary emergencies, among the policy-holders. 

 Mr. Wright takes the view that each policy earns, 

 during its continuance, an ascertainable part of that 

 reserve. He supplies the means for determining what 

 this part is, for each policy: of course, it is a matter 

 of calculation for each. He denominates this part of 

 the reserve, or surplus, the 'self-insurance.' By his 

 system it is possible tu ascertain at any time how 



much this self-insurance on a given policy amounts 

 to, or how much it will amount to at any future lime, 

 if kept in force. Mr. Wright believes that the self- 

 insurance is the property of the policy-holder; and 

 that, if not payable to him on demand, it should at 

 least be applicable for a renewal of the policy to pre- 

 vent forfeiture. 



The increase of the colored population of the 

 United States. 



BY C. S. MIXTER OF WASUINGTON, D.C. 



It is frequently asserted that the colored population 

 of the United States is increasing more rapidly now 

 than it did prior to 1861. The large apparent increase 

 shown by a comparison of the census-returns of 1880 

 with those of 1870 seems to justify this opinion; but 

 the results of investigations made by the superin- 

 tendent of census in South Carolina and Mississippi 

 show that the census in 1870 was seriously defective 

 in this respect, while that of 1880 was very full and 

 complete. The accompanying statistical table pre- 

 sents the returns of these people according to each 

 U. S. census from 1840 to 1880, and (?iTes the num- 

 bers reported from the South in det.ail. These results 

 seem to indicate that they are not increasing as 

 rapidly as formerly. The burden of supporting their 

 minor children, and their disregard of the rules of 

 health, seem to furnish additional reasons for think- 

 ing that their future rate of increase will be less than 

 it has been heretofore. 



Oyster-farming in Connecticut -waters. 



BY H. C. HOVEY OF NEW HAVEN, CONN. 



An explanation of the difference between seed 

 oysters and those fit for market gave the author oc- 

 casion to mention that ' saddle rock ' oysters in their 

 best edible condition were six or seven years old. 

 A history of Connecticut experience and legislation 

 in relation to oysters was given in detail. There 

 are now 32.5,000 acres of disposable space for oyster- 

 beds, and 100,000 are occupied. The area of the 

 natural beds is only 6,000 acres, and this furnishes 



