456 
caution the crude material might safely be used, es- 
pecially if it is not brought in direct contact a the 
plant, or if used in the dormant season. 
The experiments with phylloxerized grape-vines 
were carried on under direction of Dr. Fischer at La 
Grave d’Ambarés, near Bordeaux. Fifteen badly in- 
fested stocks, partly growing on light, partly on heavy 
soil, were qreaeed’ in April, 1888. 
It was placed in a hole dug in the ground near 
the main root, and subsequently covered up; and the 
quantity used was on some plants one, and on others 
one-half, kilogram. On Sept. 18 the plants were ex- 
amined, with the following result: all plants experi- 
mented with, but especially those treated with the 
largest quantity of naphthaline, showed a new and 
healthy growth of numerous long, fine rootlets, which 
were perfectly free from Phylloxera: in fact, the Phyl- 
loxera had entirely disappeared from the roots of all 
plants experimented with, whereas several plants not 
treated with naphthaline showed no young growth 
of rootlets, and an abundance of Phylloxera. The 
gsrowth above ground, of the plants treated, showed 
no difference as compared with plants not treated, — 
a fact explained by insufficient time for the treated 
plants to recuperate. Some of the most vigorous 
new rootlets were found to have penetrated the layer 
of naphthaline, thus showing that the latter has 
no injurious influence upon them. A considerable 
quantity of the naphthaline was found unchanged 
at the date of examination, which shows that the 
evaporation is very slow, and that its effects will be 
correspondingly Jasting. 
The results are certified to by official affidavits, aia 
were more marked on plants growing in heavier and 
moister ground than on those in light and gravelly 
soil. 
As the most convenient mode of application, Dr. 
Fischer recommends that about one kilogram of the 
naphthaline be put in a trench dug around the plant 
a few inches from the stock; the trench to be not less 
than from fifteen to twenty centimetres deep, and to 
be at once filled up again. He attributes the failure 
of former ero gone 1, to the small quantity of the 
material employed; 2, to its being employed too near 
the surface of the ground, so as to permit evaporation 
in the air. He also thinks that results were expected 
after too short a lapse of time. C. V. “Bing 
RECENT DETERMINATIONS OF STEL- 
LAR PARALLAX. 
Dr. DAVID GILL, director of the Cape observatory, 
has presented to the Royal astronomical society of 
London the results of the heliometer determinations 
of stellar parallax made by him and Dr. W. L. Elkin. 
The distances of each star, the parallax of which was 
sought from two comparison-stars situated on oppo- 
site sides of it, were measured at the times when the 
effect of parallax was least and when it was greatest. 
1 It is not stated whether the roots of these stocks were exam- 
ined at the time, to ascertain whether or not the Phylloxera 
was still at work. 
SCIENCE. 
The following were ai results obtained for the stars 
observed:— © aan 
Parallax. Probable error. 
| 
‘a Centauri: . . . .'. | 07.75 eran 
Sirius. | + 0.38 + 0.01 ° 
« Indi : | + 0 22 + 0.03 
Lacaille, 9,352 | + 0,28 a 0.02, 
oo ridanl | . Gueeeeene + 0.166 + 0.018 
2 Centauri.) } Hace oe — 0.018 + 0.019 
¢ Toucani. + 0.06 + 0.019 
e Eridani . + 0.14 + 0.020 
The probable error of a single observation by Dr. 
Gill averaged 0.1, and of a single observation of 
Dr Elkin, 0”.16. The determinations had all been 
made with the Cape heliometer of four inches aper- 
ture, and with a power of a hundred and seventy- 
five diameters. 
Dr. Gill refers to the importance of parallax inves- 
tigations in order that our knowledge of the sidereal 
system may be advanced. We do not know at pres- 
ent whether bright stars, or stars having large proper 
motions, are the more likely to give large parallaxes. 
There are, therefore, two questions to be solved,— 
first, what is the average parallax of stars of the first 
magnitude, of stars of the second magnitude, of the 
third, and so on? and, second, what connection is 
there between the parallaxes of the stars, and their 
proper motions? The present series of measures 
shows. that the parallax of a star can be deter- 
mined from sixteen measures with a probable error 
of + 0”.02, assuming that the observations were free 
from systematic errors. With a more powerful in- 
strument, which would give a greater choice of 
comparison-stars, it would seem that any systematic 
errors might be eliminated. There are sixteen stars 
of the first magnitude in the southern heavens: a 
similar number of stars might be selected of the sec- 
ond, sixteen more of the third, andsoon. In making 
these observations, a reversing-prism should always 
be employed, as in the Cape measures, that the results 
may not be affected by the position of the compari- 
son-stars. It should always be borne in mind that 
measures of two or more pairs of stars are much 
better than repeated measures of the same pair of 
comparison-stars. Another most necessary precau- 
tion is the use of screens to render the two stars 
equal in brightness. The heliometer employed should 
have a considerably greater light-gathering power 
than the Cape instrument, that there may be a freer 
choice of comparison-stars. It should be of at least 
seven inches aperture. A considerably higher power 
than the one used in the Cape determinations should 
also be employed. A single observer, by making two 
hundred or two hundred and fifty observations each 
year, might complete the entire series in the course 
of ten years. This is a work urgently demanded 
in the interest of sidereal astronomy, and one that 
should be undertaken without hesitation or delay.’ 
The heliometer of the observatory of Yale college 
is yet more powerful and perfect than that of Dr. 
a. here. 
[Von. IIL, No. 62, — 
