1903.] LOWELL—THE CARTOUCHES OF MARS. 357 
consequence was then applied to the curves of visibility as now to 
be described. 
Using the percentages of visibility as ordinates and the times 
before and after the summer solstice of the planet’s northern hemi- 
sphere as abscissze, I plotted the resulting determinations and 
connected the points so found by a smooth curve. These curves 
may be called the cartouches of the canals, since they are their dis- 
tinctive sign-manuals. Each portrays on its face the varying 
visibility of its canal during the time that it was under observation, 
but it masks much more. Were the canal intrinsically unchange- 
able, its curve or cartouche would be a straight line, since correc- 
tions for all extrinsic causes of apparent variability have already 
been applied. Its cartouche would be a line parallel to the axis of 
abscisse and at a distance from it proportionate to the canal’s 
strength. On the other hand, any intrinsic change in the canal 
reveals itself at once by a departure from a straight line. If the 
canal be for any reason augmenting, its curve will rise; if it be 
dwindling, the curve must fall. Thus the curves or cartouches tell 
us.not only of the apparent change in visibility but of the real 
change in development during examination. 
On scrutinizing the cartouches the first point noticeable is the 
well-nigh total absence of straight lines among them. There are 
but two or three instances throughout the eighty five. Thus the 
great majority of the canals were, during the time they were under 
observation, in a state of flux. For the quiescence of the remaining 
few we shall a little later in the paper be able to assign a probable 
cause. 
It is next to be noticed that opposition fell not far from the 
centre longitudinally of the curves, and the time of the planet’s 
nearest approach to the Earth still nearer the middle, since the first 
of these events happened on the 30th of March, the second on the 
3d of April. The summer solstice occurred earlier, on February 28. 
Another epoch worthy of regard is the date of the first frost in 
the Arcticregions. This, as explained elsewhere (Lowe// Odserva- 
tory Bulletin, No. 1), took place 126 days after the northern sum- 
mer solstice. It is indicated in the first diagram by a dotted line. 
On casting one’s eye down the list of cartouches arranged alpha- 
betically, no order or law is apparent. Some canals had their 
minimum early, some late, according seemingly to their own 
personal peculiarity. But if now we seek some natural order and 
