SEPTEMBER 4, 1913]| 
gave a brief account of these researches, and those 
interested in the subject will no doubt welcome the 
more detailed reports. 
Ix his presidential address to the Quekett Micro- 
scopical Club for this year, Prof. Dendy dealt with 
the subject of ‘‘ By-products of Evolution,” illustrating 
his theme by the spicules, more particularly the 
microscleres of siliceous sponges. After pointing out 
that these minute spicules exhibit constant specific 
characters, and have undoubtedly arisen by some pro- 
cess of evolution, since one form leads on to another, 
just as in the case of any other characters, it is argued 
that natural selection cannot be directly responsible 
for their origin, on the ground that the minute differ- 
ences in the form of the microscleres cannot be of any 
importance to the sponge in the soft tissues of which 
they are scattered without order or arrangement. By 
the principle of correlation non-adaptive characters of 
this kind may be linked inseparably with other cha- 
racters which being adaptive, are directly influenced 
by natural selection, in such a way that any variation 
in the one must be accompanied by a corresponding 
variation in the other. Thus, a _non-adaptive 
character may undergo a progressive evolution in- 
directly controlled by the action of natural selection. 
The principle of correlation cannot, however, be 
invoked to explain the specific forms assumed by the 
microscleres; it can only help to explain why such 
characters exist at all and why they should undergo 
progressive evolution. The specific form of the micro- 
sclere must be produced by chemical and physical 
causes involved in, and controlled by, the hereditary 
constitution of the mother-cell. 
Str F. W. Moore contributes a useful paper on 
hardy water-lilies to Irish Gardening (vol. viii. May, 
1913), including not merely cultural hints and lists of 
species suitable for ponds of different depths, but also 
some interesting remarks on the general biology and 
mode of growth of these plants. For instance, the 
author lays stress on the importance of the study of 
roots to the gardener; observation of water-lilies 
shows that from early April to June new roots are 
developed rapidly as the new leaves and flower-buds 
are formed and the rhizome elongates, while the older 
roots largely die away after having served as collectors 
of food reserves during previous years and as anchors 
during the winter. It is also noted that while, as a 
rule, the flowers close on bright days between three 
and four o’clock in the afternoon, if after noon the 
day becomes wet and gloomy the flowers usually 
remain open until dark. ; 
In continuation of his investigations into ‘‘ Southern 
Hemisphere Seasonal Correlations ” (Nature, August 
7), Mr. R. C. Mossman contributed a fourth article 
to Symons’s Meteorological Magazine for August. He 
pointed out an interesting instance of the temporary 
character of many correlations. The example chosen 
was the comparison of April to September rainfall at 
Trinidad (West Indies) with that at Azo (Argentine 
Republic) for the following six months. Dealing with 
the fifty years, 1862-1911, it was found that from 1862 
to 1877 and from 1895 to 1911 there was no relation 
between the rainfall of the six-monthly periods; but 
during the seventeen years 1878-94 the curves showing 
NO. 2288, VOL. 92] 
NATURE 
17 
the rainfall departure from normal are the reverse of 
each other. The author observes that these results 
are of importance, as they show that the physical 
processes that produced a given precipitation at Trini- 
dad during the period under discussion were associatéd 
during the six months following by an opposite effect 
in the south temperate zone, some 2850 geographical 
miles distant. Mr. Mossman also refers to one or two 
interesting correlations in other regions, especially one 
between the rainfall of Java and Trinidad. 
Tue Meteorological Office chart of the North Atlan- 
tic and Mediterranean for September (first issue) shows 
that the last report from the scout-ship Scotia was 
dated August 7 in 54° 45’ N., 49° 30’ W.; no ice in 
sight. It is pointed out that the full scope and value 
of the work accomplished cannot be estimated until 
the reports of the captain and scientific staff have been 
submitted. The ice notices which may prove to have 
been of most value are those relating to the compara- 
tively small quantities that have been seen drifting 
south in the polar current. An important feature this 
year is the fact that the ice has been held up, for the 
most part, north of latitude 43°. The special reports 
above mentioned will, it is thought, no doubt decide 
whether this was due to abnormal strength of the 
Gulf Stream, to unusual weakness of the Labrador 
current, or to both causes. 
A circuLar headed ‘*Road Dangers’ has been 
widely circulated by the editor of The Automotor 
Journal, It suggests that the dangers both of 
vehicular and foot-passenger traffic might be greatly 
minimised if at crossings the traffic of one street were 
arbitrarily given a right of way and the traffic of 
the other street which crosses it were made to go 
dead slow by a sign that must be obeyed. The writer 
of the circular considers that not only would the 
accidents that occur from collisions of vehicles at 
crossings be greatly reduced, but the noisy use of the 
horn would be no longer necessary. It is difficult to 
see how the suggestion can be carried out without 
some enactment giving to a street authority a power 
to make bye-laws controlling the traffic in the less 
important streets and which can be enforced by the 
police. It would be easy for the Chief Commissioner, 
through his advisers in Scotland Yard, to decide which 
streets are to be of primary and which are to be of 
secondary importance, but short of keeping a constable 
on traffic duty at every crossing it is difficult to see 
how, with his existing powers, he can instruct them 
to summon drivers disobeying notices informing them 
that they must give way to traffic in the preferred 
streets. 
WE have received a copy of the third edition of 
Merck’s ‘‘ Reagenzien-Verzeichnis.”” It is a volume of 
446 pages, and all the commoner reagents, tests, 
hardening and preservative fluids, and the like are 
given alphabetically under authors’ names, some 5000 
formula being thus detailed, with references to the 
literature. There is further a valuable list of the sub- 
stances for which the tests are employed, and a similar 
one for those used in microscopic work. Finally, there 
is an index of the preparations employed for the 
various tests, with authors’ names attached; thus 
’ we find that ‘‘arbutin ’ was recommended by Reichard 
