130 
the extreme ends of the framework. As the speed attain- 
able is upwards of 20 knots—developed in a 200-feet run— 
powerful emergency brakes actuated by hydraulic pres- 
sure are provided, in addition to the ordinary friction 
brakes, and the whole carriage is under perfect control 
for stopping and starting, and maintaining constant speed 
during a run. 
The extreme length of water surface in the Washington 
tank is 470 feet, of which about 370 is of the full section 
across, the remainder being the narrow extremities 
available for starting and stopping. In order to reduce 
the time lost between runs through waiting to obtain still 
water, side troughs 12 inches square in section are laid 
throughout the length of the tank to absorb the wave 
disturbance caused by a model run, while at the north 
end of the tank (Fig. 1) a series of wooden strips placed 
vertically act as a wave breaker at the close of the run. 
Great care is taken toensure the purity of the water 
by treating it with alum and filtering through sand 
Fic. 4.—Model Shaping Machine. 
before admission to the tank. The capacity of the latter 
is about one million gallons, and the tank can be pumped 
dry when required in about four hours by a 12-inch cen- 
trifugal pump. The temperature in the building is kept 
as far as possible uniform and slightly above that of an 
ordinary living room. 
A special feature at the Washington establishment is 
the employment of wood for the models instead of | 
paraffin. 
too great to allow of the latter material being used, 
preferable though it is in other respects ; for the cost of 
wood is higher, the difficulties of shaping it to the specified 
lines are greater, more time is required, and it is, of 
course, impossible to reduce it to bulk after use, as in 
the case of paraffin. On the other hand, a wooden model is 
less liable to accidental damage and retains its shape 
better if required for future use. 
NO. 1701, VOL. 66] 
This is on account of the heat in summer being | 
NATURE 
| JUNE 5. 1902 
Fig. 3 shows a model being towed through the water, 
the wave formation being clearly visible; Fig. 4 gives a 
view of the shaping machine at work on a model. 
R. W. D. 
VOLCANIC DUST FROM THE WEST INDIES. 
T was mentioned last week that the West Indian 
mails had brought packets of volcanic dust which 
fell at Barbados and elsewhere to several institutions and 
investigators in this country. The characteristics of this 
material have been minutely examined, and the following 
descriptions of them will be found of interest. 
1p 
At the meeting of the Geological Society on Wednesday, 
May 28, Dr. Flett communicated a preliminary note on the 
ash which fell at Barbados. The specimens had been for- 
warded by Dr. Morris, of the Imperial Department of Agri- 
culture, to Prof. Judd, who placed them in the hands of Dr. 
Flett for examination. The ash consists principally of 
plagioclase felspar allied to labradorite, hypersthene, monoclinic 
augite and magnetite. The crystals are often perfectly idio- 
morphic, and it may be safely inferred that they were formed in 
the magma before the actual eruption took place, and blown 
into the air along with the molten material by the force of the 
escaping gases. A small amount of glass containing steam 
holes is adherent to some of the crystals, but many are perfectly 
clean. 
The crystals are similar in every respect to the phenocrysts of 
hypersthene-augite-andesite, a type of rock well known among 
the recent volcanoes of the Pacific region. In the discussion 
which followed the reading of Dr. Flett’s paper, it was pointed 
out by Mr. Prior that the same type of rock occurs in other 
West Indian islands, and also in the Mexican volcanoes, so that 
the petrographical evidence serves to connect the West Indian 
volcanic region with the Pacific rather than with the Atlantic. 
An analysis of the ash by Dr. Pollard was communicated by 
Dr, Flett. It is quoted below. 
SiO, 52°81 MgO 519 
TiO, ‘95  K,O 60) 
Al,O3 ... 18°79, Na,O 3°23 
Fe,O3 ... 3:28 P.O; “15 
FeO 4°58 SOs 33 
MnO Ashe (Cll ue 14 
(CoNi)O ‘o7 ~=H,0 105° “20 
CaO 9°58 HO above 105° ‘I7 
Total 100°35 
It must be remembered that this analysis does not represent 
the composition of the material as it existed in the subterranean 
reservoir immediately before the eruption, but rather the bulk 
analysis of the crystals which had separated out, together with 
only a small admixture of the glass. 
If this glass could be separated and analysed it would pro-. 
bably be found to differ from the bulk analysis of the crystals 
in the same way as the glassy base of hypersthene-andesites 
differs from the bulk analysis of the phenocrysts ; that is, it 
would contain more silica, less lime, iron and magnesia, and 
more alkalis, especially potash. In Old Red Sandstone times 
the volcanoes of the Cheviot district erupted hypersthene- 
andesites, and the glassy base of one of these rocks was 
analysed by Dr. Petersen with the following result :— 
SiO, 66°25 MgO 28 
Al,O3 13°59 K,O 4°95 
Fe,O3 3°11 + Na,O 2°25 
CaO 2‘75 H,O 5°89 
Total ... 99°07” 
The samples of ash from Barbados hitherto examined con- 
sist mainly of the crystals. The glassy matter which represents 
the mother liquor appears to have been vanned away and de- 
posited elsewhere. This, if it should turn out to be the case, 
is somewhat unfortunate, for the glass, with its higher per- 
centage of potash, would have been more useful as a fertilising 
agent. J. J. H. TEALL. 
