66 
NATURE 
[ Nov. 17, 1881 
The ccrresponding numbers for the other four observers of this 
class would be substantially the same. On the other hand, Mr. 
Hart requires much more red than I do in order to convert a 
given green into yellow—in the ratio of about 2°6 : I. 
Except in the case of Mr. Hart, the colour vision of these 
observers is defective only in the sense that it differs from that of 
the majority. Their appreciation of small colour differences is 
as distinct as usual. In order to test this Mr, G. W. Balfour 
made a complete series of colour matches with revolving disks in 
the manner described by Maxwell and in my former paper, Six 
matches, of which only two are really independent, were ob- 
served, the consistency of the set being a measure of the accuracy 
of ob-ervation, The average error proved to be only double of 
that which I have found in my own observations, and rather | 
less than that usually met with in the case of observers whose 
vision is normal, 
In connection with what has been described above with respect 
to trichromic vision, it is interesting to notice that corresponding 
and perhaps larger differences are to be found in the vision of 
the so-called colour-blind. The double-refraction apparatus 
may conveniently be used in this investigation, With the 
pointer adjusted to oor 25, we have in the upper half of the 
field pure red or pure green respectively, and in the lower half 
pure yellow as usual. By suitable adjustment of the gas taps 
two observers of this class, Mr. T- and Mr. B——, are 
able to obtain perfect matches both between red and yellow, 
and between green and yellow, but the proportions necessary 
are very different for the two observers. In Mr, T- ’s red and 
yellow match, the red is to normal vision dazzlingly bright, and 
the yellow almost too dark to be recognised; while the green 
and yellow match, however extravagant as to hue, appears 
reasonable in respect of brightness. On the other hand, to Mr. 
B——’s eyes, the red of the spectrum does not look nearly so 
dark, and the equivalent red and yellow appear to the normal 
eye to be much more nearly upon a level. Although these great 
differences exist, there is no doubt that the vision of both 
observers is strictly dichromic, and that, apart from brightness, 
all the rays of the spectrum, from red to green, have the same 
effect upon their eyes. 
If we wish to go beyond the fact that this vision is dichromic, 
and inquire whether the case is one of red blindness or of green 
blindness, we must be careful to consider whether the question 
itself has a definite meaning. If trichromic vision were 
always the same, and if a particular case of colour-blind 
vision differed from it merely by the absence of the red sensa- 
tion, that vision would intelligibly be characterised as red-blind. 
There is reason to believe that such cases exist. In all proba- 
bility the suppression of my own red sensation would lead me 
to make matches very nearly the same as Mr. T——’s; and in 
this sense he may fairly be called red-blind. But under the 
same circumstances my matches would be altogether rejected by 
Mr. B 3 and the question may be asked, whether his case, 
being certainly not one of simple red-blindness, can be brought 
under the head of green-blindness. To this the sufficient answer 
isthat if I became green-blind my matches would differ from 
those of Mr. B far more than if I was red-blind. The test 
of green-blindness would be the possibility of matches between 
colours which to normal eyes appear green and purple, or green 
and grey. Although a good deal has been said lately on this 
subject, I am not aware of acase in which accurate matches of 
this kind have been obtained from observers whose colour-vision 
is in other respects acute. If such cases exist, inquiry should 
be instituted, in order to see how far the matches would corre- 
spond to green-blindness of an otherwise normal eye. 
We see, then, that there is dichromic vision which cannot 
accurately be describea as affected with red-blindness, and still 
less as affected with green-blindness. The difference from normal 
vision, being not simply one of defect, cannot be defined by any 
single phrase. To obtain a complete knowledge of it quantita- 
tive observations over the whole spectrum, such as those carried 
out Ly Maxwell, are necessary. It is fortunate that these ob- 
servations are easier to arrange for dichromic than for trichromic | ¢ 
| the substance, so that it is not easy to determine even the direc- 
vision. 
That I might be able to form an opinion upon the general 
acuteness of his colour vision, Mr. T—— was good enough 
to observe a series of five colour matches between red, white, 
blue, green, and yellow, one being left out each time. The 
results are given in the accompanying table; those marked 
“calculated ” being a consistent set derived by elimination from 
the two marked A and B, The good gereral agreement of the 
two sets of numbers is a proof that within its restricted range 
Mr. T——’s sense of colour is acute. The first observation 
in which a mixture of red and white is matched by a mixture of 
green and blue is the most characteristic. 
} 
et 
Red White. Blue. ) Green. Yellow. [Dec 2, 1880, 
| / | 
. Daal 
a} 76'2 | 23°8 —23°3 | -76°7 fe) Observed — 
774 | 22°6 -21 -79 ° |Calealated ; 
(2) | 56° | “As 523°] “0 -—47°7 | Observed 
lL! 562] 438 —52°5 ° -47°5 /Calculat 
(o{ 68°2 5°5 o | -100 | 26°3 | Observed 
2 69°77 | 65 o | -100 | 23°8 | Calculated 
(4) 60°3 omy 8 -100  31°7 | Observed 
BM) 612 ove] 7°8 | -100 31 Calculated 
{ o | 32°5 | -43°5 | 6775 | —56°5 | Observed 
Om lh S25 -441 | 677 | —55°9 | Calculated 
A | sez | 424 | -srr | 35 |= 470) 
B -470 | -199 | -377 | cx 
In conclusion I will describe an apparatus by which it is pos- 
sible to observe these colour-matches without rotating the disks. 
At the time of my first experiments, about ten years since, I was 
struck with the advantage which might ensue if it were possible 
to have the mixed colours in view during the time of actual 
adjustment, and I thought of a plan by which this object might 
be attained. The idea, which I carried out soon afterwards, 
was to spin an image of the disks instead of the disks them- 
selves, An inverting prism was mounted in a tube which could 
be made to rotate. The axis of rotation is adjusted so as to 
point accurately to the centres of the disks mounted as usual. 
An eye applied to the prism sees the disks undisplaced as a 
whole, but inverted by reflection. As the tube rotates, the 
image of the disks rotates also, and with double angular velocity. 
When the speed is sufficient, the colours lying on any circle 
concentric with the disks are blended exactly as if the disks 
themselves revolved. 
This apparatus is quite successful ; but its real advantages of 
working at a smaller velocity, and of allowing adjustment while 
the rotation continues, are counterbalanced in practice by the 
inconvenience of having to look through a tube, and the uncer- 
tainty introduced by the prs-ible disturbance of the match due 
to unequal illumination of the area occupied by the disks. 
MAGNETIC DISTURBANCES, AURORAS, AND 
EARTH CURRENTS! 
THE object of establishing a magnetic observatory is to 
determine at any instant the direction and magnitude of 
the earth’s magnetic force. The direction of the magnetic ferce 
of the earth is the direction in which a small magnetic needle 
would point when it is freely suspended, so as to turn about an 
axis passing through its centre of gravity. But it is not easy to 
suspend a magnetic needle so as to turn freely and yet to be 
sure that the axis about which it turns passes accurately through 
the centre of gravity of the needle, and if it does not so pass, 
then on suspending the needle we have not only the magnetic 
force but also the gravitating force of the earth acting upon it 
to iurn it about its axis, and the position which it takes up shows us 
the direction of these combined forces upon the magnetic needle. 
This direction depends upon the mass of the needle, for to 
that its weight is due ; it depends upon the form of the needle 
and the position of its centre of gravity with regard to the axis 
on which it is hung; it depends also on the magnetic properties of 
tion of the magnetic force by a plan which theoretically is so 
very simple. Instead of attempting to make the required deter- 
minations by such a method it is necessary that a steadier mode 
of suspension should be adopted, and that may be done as soon 
as it is discovered in what vertical plane the force of gravity, 
¥ Lecture delivered at the Royal Institution on Friday evening, June 3, 
1881, by Prof. W. Grylls Adams, F.R.S. 
