Page:Science vol. 5.djvu/471

 Mat 28, 1885.]

��tbiii, and transparent, bo that the u'liiet oi^siis with- in were visible through it. There were no eye-spots. The mouth, when expanded, appeared as an ovoid sucker, trith the orifice Bomewhat diamond-shaped; and it was neither armed with teeth, nor provided with a proboscis. The oesoiibogus is narrow, and opens Into a capacious slomoch. wtiich forma t«n or eleven horizontal discoid saccules, which were filled with a blockish-brown liquid, apparenllf blood. The stomach [s surrounded by eight tortuous, eland-like nrgaiis, which extend the e[itit« length of the body, and give rise to the colored alripes sreii through the skin. These organs are composed of numerous pyriforra acini, and appear (ike racemose glands, but their nature I did not deter- mine. The itpecimens were preserved in alchohol with the view of further Icivesti- gation, but they have softened to such a degree that the examination has proved unsatisfactory. From the conspicuous

S;land-llke organs and the habit of the eech, I propose to name it Adenobdella oricola.

In the stomach of the same flsh there were some little tkpe-wornis. which I sup- pose to be the Taenia torulusa, originally described from Eurnpeau species of Leu- ciscua and other species ut the same family. The worms were white, Glifonn, compressed cylindrical, and (rom three to six inches long. The bead is oval, with- out roatrum or hooks, and with four equi- distant, spherical, immersed bothria. The neck is narrowed and moderately long. The body widens to the posterior fourth, and tiien gradually narrows. The segments are wider than long, and not prominent. The generative aper- tures are marginal, with the penes projecting; diame- ter of the head, one-third of a line ; gi'eatest breadth of Lhe body, ttu^e-tonrths of a line.

JOSEFH LEIDY.

��Tliat f ^urea have a great capacity for lying, and that nothing needs closer watching than an argument based on statistics, are facts which ought to be well Impressed on everybody's mind. On almost every subject of public importance, — politics, finance, economic policy, social science, — one is continually solicited to believe in this or that doctrine because statistics ' prove ' It to be true. And a large part of the error that prevails on many o( these subjects — notably, on the question of free trade and protection — is due, on the one hand, to the reckless way!u which statistics are handled by writers, and, on the other, to the absence among their readers of a whole- sonie suspicion of statistical annimenis, and of the abiding consciousness that statistics do not always mean what they seem at drst sight to say.

Such being the case, it is a pity that Professor Newcomb — than whom surely no one is more free from the tnental defects to which theae errors are nsually due — should have made so many slips in a recent article in Science on mortality Matistics. One cannot help asking whether Horaer'a nods come, like mUfurtunes, many at a time.

A curious logical slip occurs In the pasaage relat- ing to the influence of occupstirin upon mortality, "How little mere occupation has lodo with viability, is ahown by the fact, that, while bankers and capital-

��ists suffer one-fourth leas, brokers, speculators, and operators sufCer twelve per cent more, Uian the tabular mortality." In other words, from the fact that in two occupations seemingly very similar the rales of mortality ore widely different, the inference Is drawn that occupation has little or nothing to do with the matter. Ohviously. the true inference is. that either the statistics are Inadequate to the making of the comparison 111 question, or that the occupations which seemed to be similar are really widely different. If we are sure the occupations are pracllcally alike, we must conclude that the statistics are iniiufflclent, or subject to a bias ; if we are sure that the statistics are sufficient and Impartial, we must conclude that some '~iportant difference is to be found In the occupations ;

��of a legitimate banker or sound capitallat.

Ill the same paragraph we are told that travelling- agents have the greatest viability of all. This Ts somewhat surprising; but the fact is deprived of all significance when one finds, on turning to the. tables, that the total number of deaths in this claas was only eight So with r^ard to the eitcessive mortality of the younger class. The whole number of deaihs b«- iweun the ages of seven and twenty Is foriy-ieven, OS Professor Newcomb mentions, while the Ameri- can table would make it thirty-three. An aggregate excess of fourteen deaths Is too slender a basis to rest any inference upon, and is not so surprisin" as to render an explanation absolutely necessary. It hap-

fens, however, that it is in a great measure explained y the fact that (as pointed out In the text accom- panying the tables) almost the entire excess occurs among the lives insured under term-po!icie»i; i.e., policies Usued to extend over a particular period only, and taken for the purpose of covering special risks.

As to the most Important point discussed by Pro- fessor Newcomb, — whether Herbert Spencer, and those who share his 'superstition,' are right In believ- ing that the most active and enterprising Americans injure their health, and shorten tbeir lives, by too great devotion to business, — 1 cannot think that these mortality statistics are any thing like a 'sura test' of the question. The class referred to Is mixed up with other classes; and, luiless we ciiii compare the mortality in this class with the mortality in ine same class in England, our inferences must be very guarded Indeed. Moreover, there are many things affect- ing selection — strictness of examination, privilege of surrender, popularity of life insurance — which may greatly differ in the two countries, and lately iuQuence the result. The great excess of raortal- ity in the case of term- policies, and the consldent- ble deficiency in the case of paid-up policies, shown by the Connecticut mutual tables, are Instances of this sort of phenomenon. And, even if we were in possession of a perfectly fair coropaiison with Eng- lishmen, it would still remain to consider whether Americans would nut, in the absence of habits com- plained of, comparestlli more favorably with English- men. On theqne&tionof Iheeffectof overwork, and worry, and ambition to become rich, a little bold a priori reasoning is likely to lead to a sounder result than can he derived from statistics not specially desii;ned to test the question. It may be remarked, as throwing some lighton the matter, that the acttuuT of the Connecticut company, after observing that be- tween the ages of fifty-six and seventy-five an undue proportion of the deaihs occur among those insured for large amounts. add». "These results ^-uggcst the

question whether those who insure for large ai

— often, jierhops generally, men of good i"

�� �