Sept. 12, 1872 J 



NATURE 



407 



lion of the terrestrial moUuscs of the Sandwich Islands seem to 

 throw light on the subject. A forest region on the island of 

 Oahu, 40 miles in length, and 5 or 6 miles in breadth, furnishes 

 about 175 species, represented by 700 or Soo varieties. The 

 average area occupied by each species is about 5 or 6 sqi-are 

 miles, though many are restricted to half that area. The valleys 

 that lie on one side of the mountain range that traverses this 

 district preserve, as far as wc can observe, the same conditions ; 

 but the varieties, and in some cases the species, found in each 

 valley, differ from those found in any other.* 



If we would account for these facts on the hypothesis of evo- 

 lution, it sf cms necessary to suppose : First, that these molluscs 

 possess an inherent tendency to variation, so strong that all that 

 is needed to ensure the divergence of type in the descendants 

 from one stock is to prtvent, through a series of generations, 

 their intermingling with each other ; and secondly ,'that either the 

 tendency to variation in this family is very much greater than 

 usual, or their tendency to migrate weaker, and their opportuni- 

 ties fewer than usual. An investigation of the conditions under 

 ■which that species exist leads me to believe that the smallness 

 of the areas occupied by each is due to freedom from that com- 

 petition that retards variation in Continental species, rather than 

 to any deficiency in the means of transportation. On the con- 

 tinents, "Natural Selection " arising from severe competition 

 with species that have a wide range, tends to prevent the develop- 

 ment of varieties, and to give a wider diffusion of forms, that 

 would otherwise be limited in their range, and variable in their 

 type. 



Mr. \A allace agreed with the Rev. J. T. Gulick in his inter- 

 pretation of facts which appeared to be exceedingly remarkable. 

 He had had the opportunity of working at a limited group of 

 organisms in a small part of the world. The results he had 

 described were a type of what took place over whole continents, 

 and exhibit an example of variation and geographical distribu- 

 tions, perhaps the most remarkable that occurs on the surface of 

 the earth, \yith the general principle that variation does not 

 depend on difference in external condition, he altogtther agreed. 

 He thought in this n-atter that there was a confusion of two 

 distinct things, even in some cases by Mr. Darwin himself 

 Variation was confounded with the formation of varieties. That 

 it was not dependent on the change of conditions was evidenced 

 by the fact that the varieties of domestic nnima's and plants were 

 not due to this cause, but only to advantage being taken of 

 spontaneous variation and identical conditions. Horticulturists 

 obtained new varieties of any plant that was introduced into cul- 

 tivation by growing it upon a very large scale, and selecting 

 the sports which were sure to occur. In this case variation was 

 accumulated by artificial selecticn, just as it is accumulated in 

 nature by natural selection. This requires, as a condition of its 

 action, a change of external conditions. We all know that 

 closely aUied, though distinct species, were found inhabiting dis- 

 tinct areas— for example, islands; and with large continental 

 areas it was the same. This had led to the very general idea that 

 it was variation of conditicn over those areas which had produced 

 the varieties, whereas it had merely selected them. In the Sand- 

 wich Islands there was no difference of physical conditions 

 adequate to produce this result. This was seen in the number of 

 intervening forms which existed. It seemed due to the absence 

 of any weeding-out effect. The land molluscs haa hardly any com- 

 petitors to struggle with, .and no enemies, quadrupeds and 

 reptiles being absent, and birds few. The rivers were small and 

 would only distribute any 'form through the same valley. All 

 these conditions favoured this remarkable persistence of closely 

 linked forms. 



SECTION G— Mechanical Science 

 On RoUing in a Semt;ay,}[iy Mr. W. Froude, F.R.S. 

 This was a description of an apparatus for automatically record- 

 ing the rolling of a ship in a seaway .and the slopes of the waves. 

 The fundamental principles on which the performance of the 

 apparatus depends are (I Mhat when waves act on a ship or 

 other floating body which would stand stably upright in stdl 

 water, she will be for the moment in equilibrium if upright or 

 normal to the mean or effective slope of the wave which she 

 occupies ; and if she possess a given righting force when inclined 

 to a given angle in still water, she will be urged by approxi- 



* A fuller statement of the fact has been given in an article 

 of Species related to tlietr Geographicil Distribution," in 

 iSlh, 187a. 



' Variation 

 ruRE July 



mately the same righting force towards the normal position in 

 wave water, if she at any moment deviate from it by the same 

 inclination. (2) A plumb line or pendulum, if its point of sus- 

 pension be at or very near the ship's centre of gravity, will be 

 lor the moment in equilibrium if it occupy the noimal position, 

 and if it have a very short period of oscillation it will instantly 

 assume that position throughout the changes of the wave slope. 

 These two propositions are but expressions of the interdepen- 

 dence which exists between the charge of translatory motion 

 which at any moment affects a mass or particle of matter, and 

 the direction at the same moment proper to any force-diicction- 

 index carried by the man, whether it be a plumb-line, which lies 

 in the direction, or a spirit level, which lies at right angles to it ; 

 the direction being simply the resultant of gravity, and of the 

 disturbing forces which at the moment affect the mass. Mr. 

 Froude described his apparatus as follows : — A revolving cylinder 

 covered with paper and turned by rough clockwork receives the 

 marks made by several pens. One of these pens records time, 

 jerks being given it at successive equal intervals by an exact 

 clock. The apparatus being placed at the centre of gravity of 

 the ship, a pendulum of very short period and considerable 

 power, oscillating in the plane transversely with the keel, records 

 continuously by a second pen the angles made at each moment 

 by the ship, with the mean or effective wave slope, that is to say, 

 her relative inclinations. Another pen, actuated by a rocking 

 arm kept level by the observer on deck, who points it to the 

 horizon, records the angle made at the same moment by the ship 

 with the horizon, that is to say, her absolute inclinations. 

 From the records thus obtained, the amount of the roll 

 of the ship with regard to the wave slope is at once shown, 

 and the form of the wave can be easily worked out graphi- 

 cally, the wave slope at each moment being simply the 

 difference between the records produced by the pendulum pen 

 and the horizon pen respectively. But the graphic integration of 

 the results supplied by the pendulum pen, if correctly peilormed, 

 supplies what might be called the theoretical measure of the 

 oscillations, which the ship ouglit to h.ave performed with regard 

 to the horizon during the period embraced in the record. For 

 the pendulum record itself supplies, throughou'^, a measure of the 

 accelerating force by which the ship's oscillation is governed ; so 

 that the integration of this gives a diagram representing the 

 angular velocity which the ship should theoretically have 

 acquired under the operation of that force. And th.e integra- 

 tion of the velocity diagram in turn gives the sequence or 

 total of motions which the v.irying velocity involves. The 

 performance of these integrations involves intleed a correct 

 knowledge of the ship's dynamic constants, but these, so far as 

 they are not already known by calculation, n".y--j)e readily 

 obtained by a single experiment with the ship in Aill water, 

 where, if she be artificially brought into oscillation (an operation 

 easily performed), and the instrument be made to record ine oscil- 

 lations as they subside under the influence of lesistance, the 

 natural period of her oscillation is at once known, and the coeffi- 

 cient of resistance is deducible in a shape which is approximately 

 applicable to the ship's seaway oscillation, All the conditions 

 required for the integration are thus supplied. Several series of 

 diagrams thus obtained by the oscillation of ships in a seaway 

 have been thus integrated, and the theoretical oscillations accord 

 so completely with the recorded oscillations that the fundamental 

 elements of the theory of rolling have been most satisfactorily 

 verified. Mr. Froude said he had more recently contrived and 

 executed an apparatus which woidd substitute an automatic 

 record of the ship's absolute inclinations for that supplied by the 

 observer on deck, as above described. For this purpose he 

 employed a heavy stationary wheel, which was so delicately 

 supported as to be incapable of receiving any rotation from the 

 motion of a ship. This wheel, if placed transversely in the ship, 

 would remain still at rest — that is to say, without rotation — and 

 would thus, while the ship performed oscillations of rotation 

 under it, communicate to one of the tracing pens a virtual 

 motion along the record cylinder, so as to form a continuous 

 record of the ship's absolute inclinations. The wheel is 3 ft. in 

 diameter and 200 lb. in weight. Through the boss is carried 

 out a strong steel axis, the prolonged ends of which are coated 

 with hardened steel. The axis thus prolonged rests between 

 two pairs of rocking arms, the ends of each pair forming a kind 

 of V. The ends of the arms arc, in fact, hardened steel plates, 

 forming segments of circles struck from the axes or centres on 

 which the arms rock, so that they are virtu.ally portions of the 

 circumferences of very large friction rollers. In order still 

 further to reduce the friction of the working parts, the axes of 



