]\gt ; On the Tathematical Theory of .329 observed discrepancy between two corresponding values are not of a very serious kind .
So that in any discussion of the variation of twenty-three pairs of organs discussed in the present paper , or o pairs of shrimp orgallS discussed in my previous communication , it may be asstlmed as at eaSt an empirical working rule that ] 's function has the same value all local races .
The question whether bhis is rigidly true will ] to be determined by fuller investigation , based on Iarger samples : but bhe value of a merely empirical expression for the relation between abnormality of one organ and that of another is very great .
It cannot be too strongly urged that the problem of animal evolution is essentially a statistical problem : that before can properly estimate the challges at present on in a race or species we , know accurately ( a ) the centage of animals which exhibit a given amount of abnormality with regard to a particular character ; ( b ) the degree of abnormality of other gans which accompanies a given abnormality of one ; ( c ) the difference between the death rate cent. in animals of different degrees of abnormality with respect to any organ ; ( d ) the abnormality of offspring in terms of the abnormality of parents , and vice versed .
These are all questions of arithmetic ; and when we know the numerical answers to these questions for a number of species we shall know the direction and the rate of change in these species at the present day\mdash ; a knowledge which is the only legitimate basis for specuIations as to their past history and future fate .
III .
" " tributions to the Mathematical Theory of Evolution By KARL ] ARSON , , Professor of Applied 11athematics , Universit .
College .
Communicated by Profcssor BICI , F.R.S. Received October 18 , 1893 .
( Abstract .
) 1 .
If a series of measurements , pbysical , , anthropological , or economical , not of the same object , but of a of objects of the same family , be made , and a be constructed by plotting up the number of times the measurements fall within a given small unit of to the , this may be nled a frequency curx .
As a rule this frequency takes the well known form of the curve of , and such a curve be termed a frequency curve .
The latter curve is bout its maximum ordinate .
however , frequency do not take the normal , and , but ) .
Such .
arise measurements ; they have been by .
Karl Pearson .
Contributions to [ Nov. 16 , Weldon , for the measurements of a organ in crabs , by Mr. Thompson , for prawns , by Mr. Bateson , for earwigs .
They occur , however , in physics , e.g. , Dr. Venn 's barometric and thermometric frequency curves ; in anthropology , e.g. , Signor Perozzo 's curves for Italian recruits , and Dr. C. Roberts ' cnrves for the eyesight of Marlborough College boys and mortality statistics ; in economics , Mr. Edgeworth 's curves of prices , and curves I have had drawn for rates of interest .
Frequency curves may , however , be abnormal and yet symmetrical .
These are much more likely to deceive even trained statistician ; such curves might arise in target practice , and would be due , for example , to firing with equal precision , but with a change of sighting at mid-firing .
2 .
Abnormal frequency curves fall into three distinct classes : .
Asymmetrical curves best represented by a point-binomial , or by its limit a continuous curve . .
Asymmetrical cnrves which are the resultant of two or more normaJ curves , with diflerent positions of axes , different areas , and different stanlard deviations\mdash ; a term used in the memoir for what corresponds in frequency curves to the error of mean square . .
Symmetrical abnol.malCUl'ves , which are compounded of two or more normal curves having coincident axes but differen areas and standard deviations , or of two normal curves with the same areas and standard deviations but different axes .
3 .
Let bs the area of any frequency curve , let the vertical hrough its centroid , or the line through its centroid perpendicular to the of measurement be drawn , and let the second , third , fourth , fifth , and sixth moments about this centroid-vertical , , and , be ascertained .
This can be done by graphical or arithmetical processes indicated in the memoir , tables being given to assist the calculation in the latter case .
Then we can treat the three classes of abnormal curves the folIowing manner:\mdash ; 4 .
Class a.\mdash ; Let the binomial corresponding to the curve be : , where probability in favour of an isolated event , probability against , and number of ibutory " " causes\ldquo ; in a single trial .
For example : the simuItaneous spinning of teetotums with black and white sides proportional respectively to and , and the total number of imes the group of is spun .
TheJl it is easy to fit this point-binomial to a frequency curve of which the centroid vertical and are known .
The soIution for this case is not discussed in the memoir , having been already dealt with by the author .
If it be to a continuous curve corresponding to the 1893 .
] the llathematical of Evolution .
asymmetrical curve , we can proceed as follows : Imagine to be large , but the ratio either small or large ; then we can obtain generalised form of the normal curve of an asymmetrical character ; its equation referred to the centroid is where stands and is the Eulerian gamma fnnction .
* Putting for an asymmetrical curve , the equation takes an indeterminate form obtained by putting , but on evaluation we have the usua ] nurmal form : Tbis generalised probability curve fits with a of accuracy a number of measurements and observations hitherto not reduced to theoretical treatment , e.g. , barometric frequency curves .
The importance of this first dissection of asymmetrical fiequency curves lies in the fact that it measures the theoretical ntlmber of contributory " " causes\ldquo ; and the odds that an element of tion will } ) positive .
The whole theory is , however , of an elementary character , and , as biological frequency curves often tend to develop a doubJe-humped character , they do not invariably fall under this cIass , and it is not dealt with at in the memoir .
5 .
Class b.\mdash ; The general theory of the dissection of a given abnormal frequency curve into components is not deaIt with , partly on account of its exceedingly great analytical difficulties , partly because there is an a priori probability that we have a mixture of only two homogcneous groups , or from the standpoint of evolution that the species will break up at first into two , lather than three or more , families .
At any rate , the dissection into two is likely to ve us either the chief components or a of the chief asymmetry of the curve .
the curve asymmetrical , it is shown that the solution of the problem is theoretically unique , but it is pointed out that in statistical practice curve is based upon a limited of measurements , and is therefore not accurately true compound of two normal groups .
A theoretical test is given to between the bebter of two or more solutions .
The method adopted for the dissection is based on equaJity of the first five moments and of the areas of the abnormal and of its two components .
This method is jnstified in the same manner as the of the * If be large , it may be taken as approximatcly wholc , and factor in brackcts is then ullity .
E.g. , claspers of eal.wigs , height of ccruits of various incos , of boys , of abitants of Doubs , \amp ; c. VOL. LIV .
2 On the xfhemotical Theorly of volution .
[ Nov. 16 , normal ve by fitting series of observations by aid of the area and the first two moments , the first moment ives the mean , and the second the error of mean sqnare ) is iustified .
The method ] eads what is termed the mental nonic , every root of which gives a roal or imaginary solution of the problem .
The best solution ; selected by the criterion that it gives the closest approach to tho given frequency curve in the value of the sixth moment .
From the nonic is deduced a quadratic for the areas of tlJe components correspondino to each solution .
If both roots of this quadratic are and positive , we have either a mixture of two eneous species , or evolution is king the homogeneous material up into two families of diffcrent , differcnt means , and different standard deviations from the mean .
If one root of the quadratic bc positive , and the other real and negative , we have evolution destroying a certain percentage round a certain mean out an initially ' and group .
Should one of the standard deviations be imaginary , we get the perccntage of anoma o aular measurement 3 in a eneous group .
6 .
Class c.\mdash ; The solution here is and depends upon the equality of the areas and of the first six moments ; for all odd moments vanish , and we have four quantities to determine , i.e. the of each group and their standard deviations .
The sollltion depends on a quadratic the areas , and the same remalks spply as to the quadratic for Class 7 .
Rules are given for detecting whether we have a mixture of two g.roups , or whether a differentiation into species of a homogeneous matex.ial is going on ; and also rilles for measuring the amount of asyznmetry which is to be considered significant .
The former rules are , briefly:\mdash ; .
Select the most asymmetrical curve out of the curves for the organs measured ; dissect it into two curves or groups by the method for Cla , .
Select the most symmctrical curve out of the curves for the measured and dissect it into two groups by the method for Class , or , if it have nificant asymmetry by the for Class again .
Then , if the dissection is possible and the sccond is not , a real evolntion is going on ; ( ) , if the dissection is possible and the second is possible , and groups give sensibly the same , we have a mixture of two and no evoIution , nnlesH the so closely allied that one must vary directly other left legs ) ; , both sections are ] , but give groups with different percentages ; we have th evolving differently at the sanlc time .
1893 .
] Mr. G. J. iments in .
333 8 .
The theory is applied to Professor 's measurements on Nnples .
It is showm that his material is absolutely homogeneous , all roots of the nonic No. 4 .
to imaginal solutions , even its real .
On the other , it is shown that { he Naples crabs are up into two diffel.ent sized families , to evoIution in their foreheads .
theory is applied to .
Thompson 's measurements of the carapace of prawns ( 1,000 measurements ) .
It is shown that we have in the measurements a very small of onlalous results , to prawns in this organ , or that there is , on the halld , a small but unstable giant population mixed tlJe n populakion .
Which of these results is to be considered the to problem can only be determined after an analysis of the frequenc } .
curves for other From the mathematical standpoint , the menloir illustrates the death roots of equations of the ninth the calculation of the nloments of IV .
'Experimens in opism.\ldquo ; By G. J. , F.R.S. Received October 2 , 1893 .
I cannot find in the literature of heliotropism that any ments have hitherto been made on the effects of interl.:lptedillumination , when the periods of illumination arc rendered as possible\mdash ; i. e. , instantaneous flashes of .
Accordingly 1 have conducted an ensive research on cliotropism , where the flashes have been caused either by means of sparks in a dark- room , or by the of a photographic shuttel placed before the plants a obscura with light or Swan burner , at a distaucc of feet on other side of shutter .
eIectric with a Wimshurst machine , ion sparks , by means of the contrivance .
the of colldensel a large induction coil copper led to a cup of mcrcury , vbere , by , means of an suitably ctuitted by , a current was closed opcned at any desired intel.vals : each was thelefore accompanied by a illiant s thick plotc glass was interposed the and the electricttl atus .
In all the experiments hel'edescl.ibed the plants emed were mustard , grown the dnrk until they had ached a of veell 1 and ) inches .
when the contrary is stitted , in all the iments c nates were by using the same pot of first half of a rati e half of the otccted from the b cap of cardboard half the pot ,

