Fuly 1, 1886] 
solids, both the hyaline substance and the granules can be seen in 
it. The hyaline substance is more soluble than are the granules, 
and is thus less commonly seen ; it is partly dissolved, partly 
swollen up intoa continuous mass ; the less swollen parts appear 
as strings or blebs. The granules in saliva vary greatly in 
appearance ; they may be very slightly swollen, and have fairly 
sharp outlines ; or they may be more swollen and run together, 
forming pale masses of various size ; occasionally in more dilute 
saliva they are just visible as pale spheres. 
Although the mucous cells are able to turn out bodily their 
products, they do not disintegrate during secretion. As the 
decrease in the interfibrillar substance takes place, there is a 
fresh formation of substance in the outer part of the cells, z.e. 
as the cell secretes it also grows. In saliva there is no evidence 
of broken-down cells, nor are nuclei present except those in 
‘salivary corpuscles,” which, as stated by Pfliiger, are leucocytes. 
Further, there is not any satisfactory proof that the demilune 
cells multiply during secretion, and give rise to mucous cells. 
During secretion there is no increase in the number of nuclei 
undergoing indirect division. As I have previously said, I hold 
the demilunes to be secreting cells of a different nature from 
that of the mucous cells. Glands with demilunes are simply 
glands in which the ‘‘albuminous” element is reduced to a 
minimum. The apparent increase in size of the demilunes, 
described by Lavdowsky as taking place in the first stage of 
secretory activity, I take to be due to the decrease in the size of 
the alveoli, so that the ordinarily flat demilunes become more 
spherical. Moreover, the demilune cells show signs of secretory 
activity. The ‘‘ young” cells described by Heidenhain and by 
Lavdowsky are chiefly altered mucous cells. 
The network of the cell consists of two parts—one in the 
cell-membrane, the other stretching from this throughout the 
cell. The peripheral network consists of very delicate fibres ; 
at some of the nodal points there are small spherical swellings. 
From lumen to basement membrane there are twelve to fifteen 
meshes. The internal network is connected with the peripheral 
network, but it appears to me to have much larger meshes. From 
basement membrane to lumen there are in the submaxillary 
gland of the dog four to six meshes, z.e. the number of meshes 
in a given direction in the cell is about half that of the number 
of granules. 
May 27.—‘‘ A General Theorem ‘in Electrostatic Induction.” 
By John Buchanan, B.Sc. 
Part I. of this paper deals with the effect of change of the 
specilic inductive capacity of a dielectric which is placed in a 
field of electric force, and it is proved that in general, under 
these circumstances, the dielectric becomes electrified. 
By translating the theorem into the language of magnetism a 
theorem in magnetic induction is obtained. 
The mathematical proof leads to an expression of the form 
‘3 dx _ 1,a°9 
A 2 Vr) 
where / denotes the rate of change of the apparent electrifica- 
tion of the dielectric with regard to the specific inductive 
capacity as independent variable ; + denotes the rate of change 
of the work done against electrical forces with regard to the 
same independent variable ; and V denotes the potential. 
The conditions that there may be no electrification of the 
dielectric are next obtained. The result is arrived at that, in 
order to have no electrification, when the specific inductive 
capacity is altered, the whole field of force must be occupied by an 
electrically homogeneous dielectric. It is then pointed out that 
the equations obtained express the effect of heterogeneity in the 
constitution of the dielectric medium. 
In Part II. the above theorem and some of the results ob- 
tained by Dr. Kerr in his experiments in ‘‘electro-optics,” are 
applied to obtain a theory of electrification by friction. 
The discussion leads to these conclusions :— 
“ Positive” liquids tend to become positively electrified by friction. 
‘* Negative ” 53 40 negatively 65 ay 
** Positive” solids 3 negatively aa “a 
“Negative ” a5 33 positively i ay 
All these conclusions are verified by the experimental results 
given in the paper. 
June 10.—‘‘ Fluted Craterless Carbons for Are Lighting.” 
By Sir James N. Douglass. 
On December 8, 1858, at the South Foreland High Light- 
house, and with the direct current magnetic machines of 
NATURE 
209 
Holmes, the first important application of the electric are light, 
as a rival to oil and gas for coast lighting, was carried out by the 
Trinity House, under the advice of Faraday. 
The carbons then used, and for several years afterwards, 
were sawn from the residuum carbon of gas retorts; they were 
square in section, 61 x 6} mm., and the mean intensity of the 
arc measured in the horizontal plane was 670 candle units, being 
17 candle units per square millimetre of cross-sectional area of 
the carbon. The crater formed at the point of the upper carbon 
of the ‘‘ Holmes” lamp was so small that no appreciable loss of 
light was found to occur, and the are proved to be very perfect 
in affording an exceptionally large vertical angle of radiant light 
for application with the optical apparatus. 
The most reliable and efficient machine that has yet been 
tried for lighthouse purposes is the large size alternate current 
magneto machines of De Meritens. The average results with 
these machines are as follows, viz. :— 
Two machines 
supplying current 
to one lamp 
48 volts. 
372 amperes. 
50 mm. diam. 
One machine 
E.M.F. svar 
Mean current~ ... ... 
Carbons (cylindrical) ... 
Diameter of crater in 
carbon 
} 
Mean intensity id) 
38 volts 
206 amperes ... 
35 mm. diam. 
13 mm. 18 mm. 
measured in the 
horizontal plane 
(candle units) .. 
Light per square 
millimetre of car- 
bon section (candle 
units) 
15,000 ... 30,000 
12 Coca Coder or 12 
It will be observed from this statement that the intensity of 
the arc in the horizontal plane per square millimetre of sectional 
area of carbon is about 35 per cent. less than it was with the 
small square carbons used by ‘‘ Holmes,” although it might 
reasonably be expected that with the improvements since 
effected in the manufacture of carbons, the efficiency of the old 
carbons would at least be maintained. The relative inefficiency 
of the large carbons used with the powerful currents now avail- 
able appears to be due (1) to the loss of a large portion of the 
most intense part of the arc which is confined within the crater of 
each carbon ; and (2) to the fluctuations in the intensity of the 
arc caused by the current passing between various points of the 
end of each carbon. 
For a new electric light installation about to be made by the 
Trinity House at St. Catharine’s Lighthouse, Isle of Wight, it 
is intended to utilise the large De Meritens machines that were 
used at the recent South Foreland experiments for determining 
the relative merits of electricity, gas, and oil as lighthouse 
illuminants. The electric light at St. Catharine’s is intended to 
be ‘‘ single-flashing ” at periods of 30 seconds. Each flash is to 
have a duration of 54 seconds, and to be followed by an eclipse 
of 244 seconds. It is intended to use one De Meritens machine 
during clear weather, and two machines whenever the atmo- 
sphere is found to be so impaired for the transmission of light 
that the flashes are not reaching their intended range. 
The defect here arose which is common to all electric flashing 
lights where a minimum and a maximum intensity of flash are 
adopted, viz. that the duration of the flashes of minimum and 
maximum intensity would vary in the ratio of the difference in 
the diameter of the carbons employed with one and two 
machines respectively, which in this case should be 50 mm. and 
35 mm., this mean difference amounting to 364 per cent. nearly. 
It is evident that such a variation in the duration of flash would 
seriously impair the distinctive character of the signal. 
It occurred to me, however, that, if carbons of a fluted cross- 
section were employed, the carbons for minimum and maximum 
intensity could be made of corresponding diameter, their sec- 
tional areas being proportioned to the minimum and maximum 
currents employed, and thus the flashes of minimum andmaximum 
intensity would have exactly the same duration. As all carbons 
for electric arc lights are now made in moulds, I saw that such a 
form would not involve any more difficulty in manufacture than 
if made cylindrical, while there would be less liability of in- 
ternal fracture occurring, as is often the case with large carbons 
in the process of drying and baking. Other advantages to be 
obtained with fluted carbons are: (1) a larger vertical angle ot 
radiant light from the arc, and with a higher coefficient of in- 
