498 Dr. F. Galton .
Note to the Memoir by unless the coefficients of variation be constant for local races , it is impossible that the coefficients of correlation can be constant for indices .
In other words , the hypothesis of the constancy for local races of correlation , and that of the constancy for local races of variation , stand on exactly the same footing .
The conclusions of this paper although applied to organic correlation are equally valid so far as concerns the use of indices in judging the correlation of either physical or economic phenomena .
It was , indeed , a difficulty arising from my discussion of personal judgments \#151 ; a spurious correlation between the judgments of different observers \#151 ; which first drew my attention to the matter .
Note , January 13 , 1897.\#151 ; The result described by Professor Pearson evidently affects the value of the correlation coefficients determined by me in Crangon and Carcinus ( ' Roy .
Soc. Proc. , ' vols .
51 and 54 ) , because I have always expressed the size of the organs measured in terms of body length .
In order to show the effect of this , I have lately performed , at Professor Pearson 's suggestion , the following experiment : It happens that my measures of Plymouth shrimps are recorded in a book , in the order in which they were measured , and therefore at random as regards carapace length or other characters .
I constructed from these records 420 " spurious " shrimps , in the following way : the total length of the first shrimp in the book was associated with the carapace length of the tenth shrimp and the " post-spinous length " of the twentieth , and so throughout .
Evidently these three measures were associated at random , and we might expect that these spurious shrimps would show no organic correlation ; but when the carapace lengths and " post-spinous lengths " of these spurious shrimps were divided by the body length , and the correlation between the resulting indices was determined , the value of was found to be 0'38 , the value for real shrimps being 0-81 , or the correlation due to the use of indices forms 47 per cent , of the observed value .
W. E. R. Weldon .
" Note to the Memoir by Professor Karl Pearson , F.R.S. , on Spurious Correlation .
" By Francis Galton , F.R.S. Received January 4 , -\#151 ; Read February 18 , 1897 .
I send this note to serve as a kind of appendix to the memoir of Professor K. Pearson , believing that it may be useful in enabling others to realise the genesis of spurious correlation .
It is important though rather difficult to do so , because the results arrived at in the memoir , which are of serious interest to practical statisticians , have at first sight a somewhat paradoxical appearance .
Prof. Karl Pearson on Spurious Correlation .
499 The diagrams show how a table of frequency of the various combinations of two independent and normal variables may be changed into one of A/ C , B/ C , where 0 is also an independent and normal variable in respect to its intrinsic qualities , but subjected to the condition that the same value of C is to be used as the divisor of both members of the same couplet of A and B. In short , that the couplets shall always be of the form A/ C " , B/ C\#171 ; , and never that of A/ C " B/ C* .
For the sake of clearness , the simplest possible suppositions , that are at the same time serviceable , will be made in regard to the particular case illustrated by the diagrams , namely , that A , B , and C , severally , are sharply divided into three , and only into three , equal grades of magnitude , distinguished as AI , All , AIII ; BI , BIT , Bill ; and Cl , CII , CIII ; also that the frequency with which these three grades occur is expressed by the three terms of the binomial ( l + l)2 .
Consequently there is one occurrence of I to two occurrences of II and to one occurrence of III .
Roman and italic figures are here used to keep the distinction clear between magnitudes and frequencies .
It will be easily gathered as we proceed , without the need of special explanation , that the smallness of the value of the binomial index has no influence either on the general character of the operation or on its general result .
The large figures in the outlined square , occupying the lower right hand portion of fig. 1 , show the distribution of frequency of the various combinations of A and B. The scales running along the top and down the left side of the figure , which are there assigned to the " values of A/ G , B/ C , \amp ; pply to these entries also .
The latter run in the same way as those in Table I below , or when quadrupled , as they will be for purposes immediately to be explained , as in Table II .
Table I. Table II .
12 1 2 4 2 12 1 4 8 4 8 16 8 4 8 4 Let us now follow the fortunes of one of the large figures in fig. 1 , say that which refers to A = I , B = III , of which the frequency is only 1 .
When the latter is expanded into the three possible values of the form A/ C , B/ C , caused by the three varieties of C , it yields i case of frequency to ( I/ I , III/ I ) , f case to ( I/ ll , III/ II ) , and i case to ( Hill , III/ III ) , for entry at the intersections of the lines ( I , III ) , ( I/ ll , III/ II ) , and ( I/ III , I ) respectively .
But , in order to avoid the inconvenience of quarter values , it is better to suppose the original figures in the fig. and in Table I above to have been replaced by those in Table II ; then the original entry 2 p 2 Prof. Karl Pearson on Spurious Correlation .
501 from which we start will have become four , to he expanded into three derivative entries , having respectively the frequencies 1 , 2 , and 1 ; these latter figures are entered in fig. 1 at the intersections of the lines just named .
Under this arrangement the large figure from which we started , which had been changed from 1 to 4 , again assumes its original value of 1 .
It will easily be understood , that the positions of the three derivative entries necessarily lie in the same straight line , and that this line necessarily runs towards the ( 0 , 0 ) corner of the figure .
The same is true for every other set of derivative entries , with the result that whereas the original set of large figures , referring to the combinations of A and B , are symmetrically disposed on either side of the horizontal , of the vertical , and of the diagonal lines passing through their common centre at ( II , II ) , the derivative values of A/ C , B/ C are disposed symmetrically only in respect to the diagonal line that runs from the ( 0 , 0 ) corner .
Their symmetry , in this sense , is well shown by the dotted connections between the corresponding figures on either side of the diagonal .
Also , it will be seen that the diagonal passes through the regions of greatest frequency .
It follows that the diagonal in question represents the locus of average frequency .
Now , along that diagonal , each value of A/ C is associated with identically the same value of B/ C ; in other words , a correlation is found to have become established between them , which is solely due to the fact that each member in every couplet of A/ C , B/ C values is divided by the same value of the variable C. We will now submit the above process to the te3t of extreme cases .
First , let the variability of A be so small that it may be treated as a constant , and take it = 1 .
Then the values of A/ C and B/ C , that are sevei*ally associated with the three values of C , are as follows:\#151 ; Table III .
c. A/ C. B/ C. | | Corresponding j frequencies .
j I I I II IIL - 12 1 II I/ II I/ II I min 12 1 III I/ III I/ III II/ III i 12 1 These frequencies are laid down at their pi'oper places in fig. 2 , where the three entries , corresponding to each successive value of A/ C , run in vertical lines , but , on connecting the entries of maximum 502 Prof. W. J. Sollas .
Report to the Committee appointed frequency it is seen that they coincide with the diagonal from the O/ O corner ; also that the entries of minimum frequency are disposed symmetrically on either side of that diagonal and converge towards the same corner .
Consequently , the existence of spurious correlation is manifest here .
If B he the constant , and A and 0 the variables , the general results will of course be the same .
Secondly , let both A and B be constant and equal to I , and C the only variable ; then there are only three possible combinations of A/ C and B/ C. In one of them both values are equal to I , in another to I/ II , and in the third to I/ III , all of which lie along the diagonal from ( 0 , O ) , and thus testify to intimate correlation .
Lastly , let C be the only constant and equal to 1 .
Then A/ C , B/ C , become A and B , and the table of frequency of their various combinations is that shown in Table I and by the large figures in fig. 1 , whose symmetrical disposition in all directions proves that there is no correlation .
" Report to the Committee of the Royal Society appointed to Investigate the Structure of a Coral Reef by Boring .
' ' By W. J. Sollas , D.Sc .
, F.R.S. , Professor of Geology in the University of Dublin .
Received December 31 , 1896 , \#151 ; Read February 11 , 1897 .
Prefatory Note by Professor T. G. Bonney , D.Sc .
, LL. D. , F.Ii .
S. , Vice-Chairman of the Committee .
In presenting , as desired by the Committee , Professor Sollas 's report-on the attempts to ascertain , by boring , the structure of the atoll of Funafuti and on other investigations simultaneously undertaken , I avail myself of the opportunity of expressing the gratitude which is felt by its members to our friends in New South Wales , who have given such real and substantial help , especially by the loan of machinery and skilled wmrkmen , in putting the project into execution ; and among ' them chiefly to Professor Anderson Stuart ( who has been practically another secretary in Australia ) , Professor Edgeworth David , Mr. W. H. J. Slay ( Chief Inspector of Mines ) , and Sir Saul Samuel ( the Agent-General of the Colony in England ) .
I shall venture also to acknowledge gratefully the services of Captain Field and the officers of H.M.S. ** Penguin,5 ' and the unstinted labour which has been given by Mr. W. W. Watts , F.G.S. , our Secretary in London , in carrying out our plans .
In conclusion , may I express , speaking for myself , my earnest hope that another attempt will be made to determine the true structure of an atoll .
I think , however , that our experience on this occasion shows that the attempt can be

