294 



NATURE 



[July 28, 1892 



of death, I argue thus : — Since every event ( = change of attri- 

 butes in subjects) has a cause, the death in question had a cause ; 

 the only precedent event that was relevant, was the administra- 

 tion of arsenic, therefore the arsenic was in this case the cause of 

 death (this last result is obtained by the Method of Difference- 

 by it we prove cause — i.e. interdependence of successive events). 

 But (by the principle of uniformity) if arsenic is on one occasion 

 cause of death, it is always cause of death; therefore arsenic is 

 always a cause of death. 



It will be observed that in this second induction, though not 

 in the first, we make use of one of Mill's " Inductive Methods." 

 The function of these Methods is to prove interdependence be- 

 tween phenomena— whether it be an interdependence of con- 

 comitance or of causation. In the case of the Method of Differ- 

 ence we proceed on the assumption that if the introduction of A 

 is followed by the appearance of C, or the removal of A by the 

 disappearance of C, then A and C are causally interdependent. 

 In the Method of Agreement we proceed on the assumption that 

 if A is never found without C, A has a connection of inter- 

 dependence with C. 



We do not use, and do not need, these Methods in mathe- 

 matical generalizations, because there we see the interdependence 

 upon which generalization to unknown cases is based ; it is this 

 actual apprehension of interdependence that both makes the 

 methods unnecessary and gives mathematical generalizations 

 the peculiar certainty which is generally attributed to them. 

 In the case above cited, for instance, we see that equality of 

 angles at the base is self-evidently and necessarily bound up 

 with equality of sides in a triangle. We do not see that there is 

 a self-evident interdependence between the obvious properties 

 of arsenic and poisonousness. 



A further interesting point is that our power of predicting 

 that one event. A, will be followed by another event, C, seems 

 to depend wholly upon coexistence of attributes in the subjects 

 concerned. If we have seen one animal dosed with arsenic and 

 subsequently die, and hence conclude that another animal called 

 by the same name, and dosed with an equal amount of arsenic, 

 will die, is not our inference based upon the assumption of a 

 certain constant coinherence of attributes, both in the animal 

 and in the poison— a coinherence of such a kind that when the 

 two subjects are so collocated as to act upon each other, a result 

 similar to that produced in the first case will be produced in the 

 second also ? If the properties of this arsenic are different from 

 the other, or if the second animal, though looking like the first, 

 has a different internal constitution, there is no reason why 

 death should result. Hence, laws of succession in events seem 

 to depend upon laws of coexistence of attributes in subjects. 



Even those generally unquestioned axioms of logic, the Law 

 of Contradiction and the Law of Excluded Middle, might be 

 appealed to (if it were necessary) in support of the Principle of 

 Interdependence — for the Law of Excluded Middle intimates a 

 thoroughgoing connection (positive or negative) between all 

 nameable things ; and the Law of Contradiction asserts a certain 

 definite amount of necessary interdependence of properties in 

 every imaginable case — interdependence, namely, between the 

 ■presence of any characteristic and the absence of its negative. 



Looking at the whole process of inductive reasoning, it appears 

 to be in the application of the "Methods" that the principles 

 used approach nearest to the character of mere assumptions ; 

 and this is so only because of the difficulty of applying the 

 Methods precisely— of being sure, e.g. in the case of the arsenic, 

 that the administration of arsenic was the only new antecedent 

 relevant to death. 



It may just be noticed that in an argument by analogy we rely 

 upon an mterdependence which is inferred from the complexity 

 or amount of interdependence already known or supposed. 



If the above account of inductive reasoning is accepted, it 

 appears that the connection between Induction and Deduction is 

 very close — in fact, that the one distinctive feature of logical 

 induction is the element of hypothesis or discovery — the suppo- 

 sition of a given connection — from which every Induction must 

 set out. 



Cambridge. E. E. Constance Jones. 



The Scale for Measurement of Gas Pressures. 

 I VENTURE to ask you to print the following suggestion. It is 

 one likely enough to have been made before, but I do not re- 

 member having met with it. 



NO. T 187. VOL. 46] 



We generally measure gaseous pressures in millimetres of 

 mercury, and 760 mm. is adopted as the standard pressure. It 

 would certainly be more convenient if we expressed the measure- 

 ment in degrees, the degree being of such magnitude that the 

 standard pressure were 273°. All calculations involving change 

 of P, T, and V to or from the standard conditions would be 

 simplified in an obvious way. The equation PV = RT would 

 become V=R at standard pressure and temperature. R being 

 the same constant for all gases under all conditions, if V stand 

 for the molecular volume, it would be convenient to remember 

 it as identical with the well-known number expressing the stan- 

 dard volume of a gramme-molecule. i°P would correspond to 

 about 278 mm. or \ inch of mercury. 



Orme Masson. 



The University of Melbourne, June 21. 



Luminous Clouds. 



Bright luminous clouds were seen here on the night of Sunday 

 the 24th inst., in the north and north-north-east, from 9.35 to 

 10.3s P-™- As usual they distinctly resembled cirri, having some 

 definite upward curls. The actual cirri, which had after sunset 

 been moving rapidly from east-south-east, now appeared dusky 

 against the twilight glow. The filature of the upper or luminous 

 cirri was, as appears to be usual, west and east, while that of the 

 ordinary cirri was east-south-east and west- north-west. 



These luminous clouds, although no doubt simply reflecting 

 solar light, generally appear to the casual observer as incan- 

 descent or self-luminous. 



_ They were seen from the summit of Ben Nevis all through the 

 night of the 24th-25th, according to the report in the Ti?iies, 



Lutterworth. W. Clement Ley. 



Whirlwinds in the South Indian Ocean. 



The following account of whirlwinds met with in the South 

 Indian Ocean at the end of last May, which has been supplied 

 to the Meteorological Office by Messrs. Sandbach, Tinne & Co. , 

 of Liverpool, may be of interest to your readers, 



Robert H. Scott, 

 July 22. Secretary, Meteorological Office. 



Extract from a Letter received from Cap!. S. P. Hearn, 

 Ship " Genista." 



"At noon on May 26, lat. 42° o' S., long. 99° o' E., wind 

 fresh from N.W. — weather very squally with rain, barometer 

 steady at 29'82 in., thermometer 49° since midnight. A very 

 heavy black squall with rain began to rise in the W. Barometer 

 suddenly fell o'l in. As the squall neared the ship it arched 

 up in the centre, showing a very bright blue sky at the back of 

 it ; the ends of the squall on either side were quite black and 

 thick with rain. On its nearer approach to the ship I saw two 

 immense whirlwinds, just a little on either side of the centre of 

 the arch and coming direct for the ship, the sea under and near 

 the whirls being carried around and up in great volumes. I 

 thought at first they were two waterspouts forming, but I saw 

 no descending column or clouds from above, as is seen when 

 a waterspout is forming ; when these whirls came to within two 

 miles of the ship, the squall seemed to part in the centre of the 

 arch — one half passing to the N.E., the other half to the 

 S.E., one whirl following in rear of each part of the squall, 

 and not where the clouds were heaviest. During the time of 

 the separation of the arch we had the wind very unsteady from 

 N.W. to S. W. There was only a fresh breeze with thick 

 rain in that part of the squall that neared the ship ; yet the 

 squall was travelling along at a great rate, the whirls keeping 

 in the rear till out of sight. I shortened sail to topsails as soon 

 as I saw the squall rising. After it passed, the weather looked 

 very fine, bright, and clear, but the sky was a windy one, being 

 a very bright blue. By 3 p.m. the wind shifted to W., and 

 barometer had fallen to 29-67 in., thermometer 48°. At 4 p.m. 

 saw another whirl passing along to windward in the rear of a 

 squall, the clouds above it being twined and twisted every way. 

 During the whole night we had very heavy squalls, sometimes 

 following one another very quick, with little wind between — 

 direction W. S.W. At daylight the weather was much finer. 

 After that, to lat. 40° 22' S., long. 125° E., I had very peculiar 



