280 
NATURE 
[ JANUARY 21, 1897 
blance between these two comets is greater than was at first 
supposed, the origin of the unsatisfactory large differences for 
the mean places having been found out. The elliptic elements, 
which he has now calculated, give us less reason, then, to doubt 
the probability of a connection between these two comets. Dr. 
Ristenpart compares his elements with those of comet Biela at 
the time of its appearance in 1852, but suggests that more 
observations must be used in the investigation before an accurate 
value of the eccentricity, and therefore of the period, can be 
obtained. The comparison is as follows :— 
Comet Perrine 1896. 
Comet Biela 1852. 
7 Nov. 24°7433 B.M.T. 
° ‘ “ ° ’ 
w 163 57 30 5) 223 17 
Q 246 24 7°2/1897'0 245 51 
2 13 50 4I°I 12 33 
log g  0°046412 9°9348 
loge 9843395 9°8784 
Zi 3676 37526 
U 7047 years 6°62 years 
DouBLeE Star MEASuRES.—Mr. R. G. Aitken communi- 
cates to the Astv, WMachr., No. 3395-6, his measurements of 
double stars during 1895-6 with the 12-inch and 36-inch 
equatorial telescopes of the Lick Observatory. The majority 
of the measures were made with the former instrument, but 
occasionally the 36-inch was used for any crucial test. The 
selection of objects was restricted ; no special search was made 
for new doubles, and great care was taken to determine the 
proper quadrant when two stars of nearly equal magnitude were 
being observed. In the micrometric measurements published, 
Mr. Aitken gives double weight for observations made with the 
36-inch ;'the position angle is the mean of four or more settings, 
and the distance that of three double-distances. 
The following are some of his remarks on interesting doubles 
and questionable doubles :— 
O= 65 (Mag. 6°5, 7’0).—Certainly a physical pair, and the 
plane of the orbit appears to be in the-line of sight. 
Further measures are needed at short intervals. 
H VI to1 6 Tauri (Mag. 4'0, 9°0).—Distance appears to be 
slowly increasing. 
= 634 (Mag. 5°0, 8+).—Rectilinear motion. Stars are moving 
in nearly opposite diréctions. Distance in 1834 was 34”, 
in 1896, 14°75. 
H 1222.—Examined this star with the 36-inch powers to 
1000. Star apparently single (1896°475). Conditions 
good, ‘‘ Strongly suspected close double” by H. Looked 
for by 8 in 1876 without success. Probably a mistake on 
the part of H. 
OX 269 (Mag. 6°5, 7°0).—Companion of this rapid binary has 
completed more than one revolution since the measures 
of OS. From measures down to 1891, 6 finds a period 
of 48-4 years. 
= 2026 (Mag. 8'9, 9:0).—Undoubted binary. Angular motion 
should now become more rapid. 
OZ 342. 72 Ophiuchi.—With 36-inch powers 
Apparently single (1896°488). Powers to 2600. 
certainty of elongation ” (1896°513). 
Measured as a close pair by O3 and others, but B has 
always found it single in the last twenty years. Probably 
the companion is an illusion. 
B 989 x Pegasi (Mag. 5°0, 5+).—Shortest period of any 
known binary 11°37 years. 
THE CANALS OF MArs.—We have received a communica- 
tion from Herr M. Teoperberg, of the Hague, in which he 
submits an explanation of the formation and doubling of the 
canals on Mars. The idea which he suggests is one that will 
scarcely recommend itself to astronomers, for, indeed, ove 
assumption cannot reasonably be admitted. The writer sup- 
poses a periodical downfall of snow to be the principal agent, 
taking the undoubted bands as the crests of anticlinals, the bases 
of which may be veiled from the observer by increase of 
absorption. Such a range, he says, presenting itself as a 
narrow band, will be doubled if the higher part of the crest 
be covered with snow. With the advance of the season the 
snow-covering will extend downwards on the slopes of the 
ridge, and its margin will at last dip into those strata which 
escape our observation: the bands will then be lost for a 
time, reappearing by the inverse process at the next change 
of season. As another instance of such combinations, he says, 
NO. 1421, VOL. 55] 
1600, 
oe No 
“fa synclinal, filled up in winter with snow extending also, 
but in thinner layers, over the bordering ranges, will present a 
double band as soon as these more exposed ranges are laid bare 
by the melting of the snow in summer. They will then change 
into a single band when the central thicker mass of snow has 
melted away and replaced either by the dry valley ground or by 
a drowned 7ha/weg, these recalling the canals prop. dict., differ- 
ing, however, therefrom by a probably high situation and by the 
elevated ranges on the sides.’’ Sufficient, however, has been 
quoted to show that the writer must assume in his hypothesis 
innumerable ranges of mountains, the highest peaks of which 
must be singularly placed to give the effect of straight lines or 
arcs. It is true that horizontal sections of mountains become 
more simplified the greater the elevation, and that gaps of 
considerable magnitude would escape observation, but even 
then the mountainous conditions on Mars would be very 
extraordinary. If such were the case, the ‘‘flashings” would 
be very much more numerous than they are, and the colour 
phenomena would probably be different from what observations 
tell us. The hypothesis of ‘‘ vegetation’ seems still to be the 
most satisfactory explanation for these curious canal-like 
markings, although even this cannot satisfy all the observed 
phenomena. 
THE CLASSIFICATION OF MADREPORARIA. 
At present the classification of Madreporarian corals is ad- 
mitted to be in an unsatisfactory condition. <A fair 
standard of the opinion of the time can be obtained by reference 
to Prof. Nicholson’s ‘* Manual of Palzontology,” or to Prof. 
von Zittel’s new ‘* Student’s Text-book of Palzeontology.” One 
of the most striking features is the insecurity which is now felt 
about the sub-orders of Milne Edwards and Haime, the M. 
Rugosa = Tetracoralla (Haeckel), and the M. Aporosa and Per- 
forata = Hexacoralla (Haeckel). Yet the authorities just named 
think it best to maintain these sub-orders provisionally. 
I propose here to give a short sketch of some work of mine 
on corals, which was entered at the Royal Society of London 
in July 1895. It deals with the ‘‘ Microscopic and Systematic 
Study of Madreporaria,” and covers a rather wide field :— 
(1) Numerous sections of the skeleton of living corals are 
examined and figured, all controversial points with regard to 
the structure of skeletal parts are exhaustively discussed, and 
my own views are advanced regarding microscopic structure, 
and the relation of the soft parts of the polyp to the skeleton. 
(2) A comparative account is given of the fossil skeleton in 
the various families. In this part I have been greatly aided by 
the results of my special work on corals of Upper Jurassic age, 
**Stramberger Korallen,” to be published next month in the 
Palidontologische Mitthetlungen (Stuttgart). 
(3) The determination of the main evolutionary changes within 
Madreporaria. 
(4) Systematic results. 
General Microscopy of the Skeleton. 
The basal “tabula” or ‘‘dissepiment”’ forming the floor of 
the calyx presents us with the simplest form of septal structure 
in Madreporarian corals. Microscopically examined, it proves 
to be a compact series of calcareous lamella, each of which is 
made up of minute, crystalline, needle-shaped fibres, set per- 
pendicularly to the lamellar surface. The fibro-crystals are 
oriented in one and the same direction throughout the whole 
thickness of lamellze, hence they behave in the same way to- 
wards light. The appearance under the microscope is that of 
long fibres running through the lamella, but crossed by a dark 
and light band in each lamella (Fig. 1). These bands are 
wavy, not straight in section, and a special group of tiny fibres 
is present within each ‘‘ wave.” The solid dimensions of such 
a fibrous wave-unit or ‘‘scale”’ (Fig. 2) of the lamella agrees 
with those of a single ectodermal cell of the polyp. The wave- 
units of the lamellar surface indicate the original cell units, 
whose protoplasmic contents have been changed to crystalline 
fibres. Still, however, some fragments of organic matter and 
dirt particles usually remain, and their decomposition gives rise 
to the dark spots and bands which blur the crystalline deposit. 
The structure just described for the calcareous floor of the 
coral calyx repeats itself throughout all parts of the skeletal 
