156 



KNOWLEDGE. 



[July 1, 1896. 



Here we have eight comets introduced into the system 

 in the short space of six years — on the avorap;e, more than 

 one annually — and yet the total number of elliptic comets 

 of approximately this period is only about thirty, including 

 those that are lost and some whose periodicity is very 

 uncertain, owing to defective observations made in the 

 last century. Evidently, if this rate of increase was only 

 approximately correct, the number now known should be 

 far greater. Some in this catalogue, it is true, have been 

 very faint objects, and would probably have passed un- 

 detected in days when telescopes were smaller or searchers 

 fewer in number. Still, as pointed out in my article in 

 the February issue, this suggestion does not altogether 

 explain the frequency with which elliptic comets are 

 now met with. Another very obvious suggestion presents 

 itself. Is it possible that these objects are not all 

 uidependent ? May not some of the older and lost 

 comets take upon themselves new shapes — move along 

 strange and unrecognized paths under the influence of 

 planetary perturbation, exerted upon them at some 

 period in their career subsequent to our observation of 

 them ? 



In this question we have implicitly two suggestions : 

 the possibility of identity and the possibility of intimate 

 connection between comets, arising presumably from 

 having sprung originally from one parent stock ; for it 

 is necessary to bear in mind that other causes than 

 planetary perturbation may be operative. We have seen 

 Biela split itself into two sections, and the dichotomized 

 comet pursue a bifurcated path. It may not be pos- 

 sible to demonstrate the cause of this disruption with 

 certainty, though the behaviour of other comets, par- 

 ticularly the perplexing and inexplicable changes in 

 brilliancy which they present to us at successive returns, 

 and even while under observation, suggests that it was 

 more probably due to some mechanical agency within 

 itself than to any external force. In this particular case 

 of Biela the two separated portions kept near each other, 

 and came up to perihelion together within a day ; but had 

 the operating cause been greater, it is easy to imagine that 

 a longer space of time would have separated the return of 

 the two fragments, which, pursuing paths wider asunder, 

 would have been submitted to different effects of perturba- 

 tion, and the other elements of the orbit would have 

 differed proportionately. Or suppose, the disruptive force 

 being the same, several revolutions had been completed 

 without the possibihty of observing the comet ; the dis- 

 connected portions would have separated more and more 

 at each successive return, till it is quite possible only one 

 of the components would have been seen. In the great 

 comets of 1843 I., 1880 I., and 1882 II., we appear to 

 have an instance of more violent disruption, in which one 

 may suggest that some giant comet, carrying in itself greater 

 forces, and consequently greater opportunities for producing 

 catastrophe, has so violently dismembered itself that its 

 several portions arrive at perihelion separated by years. 

 The history of cometary astronomy contains so many sur- 

 prises that it is almost impossible to reject any hypothesis 

 as absolutely untenable, and it is unfortunately too easy 

 to throw out suggestions that it is impossible to put to the 

 test of experiment or efficient analysis. For instance, it 

 is generally accepted now, with more or less certainty, that 

 comet tails are due to a form of energy (electrical repulsion 

 is the usual form the explanation takes) acting from the 

 sun in a sense contrary to that of gravitation. But 

 recent photographs of comet tails have indicated a 

 shattermg and discontinuity incompatible with a force 

 regularly and contmuously operative ; and though this 

 irregular structure may be explained by actual encounter 



with some asteroid, it is not impossible that the origin of 

 disruption was present at the birth of the tail itself. 



Theonlypointtliat is heic insisted upon is this: that while 

 planetary perturbation, acting under laws that are perfectly 

 understood and can be submitted to mathematical analysis 

 or arithmclical calculatitm, is undoubtedly a great factor, 

 either in the introduction of fresh cometary matter into 

 the solar system or in controlling cometary matter tbat 

 entered the solar system with the velocity due to parabolic, 

 or approximately parabolic, motion, it is not necessarily 

 the only force that can be invoked or employed as a work- 

 ing hypothesis. The suggestion of hypotheses without 

 submitting them to adequate test is, perhaps, a confession 

 of ignorance and an attempt to hide it ; but the criteria 

 applied by astronomers and miithematicians in this par- 

 ticular connection rarely give a certain and unambiguous 

 result. The conditions of the problem are so varied, the 

 opportunities of escape so numerous, the labour of absolute 

 test so onerous, that we are too often obliged to accept a 

 half answer, and console ourselves with the hope that time 

 will light in our favour by adding fresh observations and 

 increasing the accuracy of the deductions. But time fre- 

 quently introduces a fresh and complicating factor, without 

 by any means removing the old perplexities. 



As a good instance of the difficulties to be overcome in 

 answering the question of a comet's origin, we may take 

 the first item on the list (1891 V. Swift) and endeavour to 

 understand the interesting problem it offers for solution and 

 the means it supplies for that solution. On November 21st, 

 1894, Mr. E. Swift, the son of Prof. Lewis Swift, well and 

 honourably known in connection with cometary discovery, 

 picked up a faint comet at considerable southern declination. 

 It had passed its perihelion and was growing fainter, but was 

 moving northward, which slightly improved its chances of 

 observation, and Prof. Barnard was able to follow it with 

 the large refractor of the Lick Observatory till January 

 25th. We have, therefore, two months' intermittent 

 observation of a faint nebulous patch by a few observers, 

 no two of whom are likely to agree on the same precise 

 point for measurement, to say nothing of the possibility 

 of an alteration of shape. Yet out of these observations, 

 which cover so small an arc of an ellipse which requires 

 six years to complete its revolution, the entire curve has 

 to be constructed. Such procedure puts the c.r pede 

 Ueiviilem principle to its fullest stretch. Nevertheless, 

 as soon as a few observations permitted the first rough 

 determination of an orbit, M. Schulhof announced unhesi- 

 tatingly that it was the lost comet of De Vico, that had 

 previously been seen once, and once only, in 1844. If 

 this conjecture be well founded, evidently the comet has no 

 right to a separate place on the list, any more than any 

 other comet which may have been missed at one or several 

 returns and subsequently recovered, of which examples were 

 given in February. On what grounds, then, had M. Schulhof 

 based his assertion '.' 



The observations of the comet of 1844 had been submitted 

 to a very critical discussion by the late Dr. Brunnow, and 

 though the orbit was to some small extent uncertain the 

 elements were fairly trustworthy, and resembled those of 

 the parabola that Schulhof had computed from the first 

 observations that had come to hand. It may be as well 

 to mention here that this is the course uniformly pursued 

 by computers. A parabolic orbit is first derived because 

 the labour of computing elliptic elements is far greater, 

 and to obtain accuracy comparable with that labour the. 

 elements should be based on observations extending over 

 a longer period than is required to give lair accuracy in 

 the case of parabolic motion. So complete was the 

 resemblance between the new parabola and the old ellipse 



