—_— se... 
OcToBER 19, 1916] 
NATURE 
135 
dence, which seems to be beyond dispute, is produced 
to show that the Phoenicians played the part of dis- 
tributing agents of this culture, and not that of mis- 
sionaries. The search for wealth was the underlying 
factor in this distribution, and provided the motive for 
the widespread travellings of these people. To assist 
in this search they employed expert gold-miners from 
the Black Sea and from Colchis, hence the megalithic 
monuments in the mining camps. That these two 
contributions will give rise to much discussion in the 
immediate future need scarcely be said, for the theme 
is one of first-rate importance; but there can be little 
doubt that the views they have so skilfully expressed 
will meet with general acceptance. 
Sporapic migrations of butterflies, moths, and 
dragonflies have long been Iknown, and have generally 
been regarded as comparable to the similar migrations 
of locusts. But Mr. Howard J. Shannon, in the Scien- 
tific Monthly for September, under the title ‘* Insect’ 
Migrations as Related to Those of Birds,’’ has mar- 
‘shalled a host of facts which apparently show that 
certain North American species of butterflies, diptera, 
and dragonflies annually migrate southwards in the 
autumn, following well-defined routes corresponding to 
those taken by the birds of the same regions. ‘Lhe 
author does no more than hint at a return migration 
in the spring, expressing curiosity as to whether such 
migrants are of the same individuals which passed 
southwards in the autumn, or whether they represent 
a new generation, bred in the southern winter quar- 
ters. Some idea of the magnitude of these migrations 
may be gathered from the author’s statements in 
regard to the monarch butterflies (Danais archippus), 
which, “in mingled myriads, move forwards . . . in 
swarms... forming veritable crimson clouds... 
miles in width and streaming backwards for equal 
distances . . . casting below them as they go per- 
ceptible shadows.” 
Unper the modest title, ‘‘ Notes on Some Palzeozoic 
Fishes,'’ Messrs. D. M. S. Watson and Henry Day 
have enriched the Memoirs and Proceedings of the 
Manchester Literary and Philosophical Society with a 
contribution of first-rate importance. They have de- 
scribed and figured in detail the skulls of Holoptychius, 
Glyptopomus, and Osteolepis, and the whole fish of 
Rhizodopsis. | These studies have shown that the 
Rhipidistian skull is considerably more complicated 
than has previously been supposed, and that the short 
type of paired fin found in Eusthenopteron is older 
than, and has given rise to, the biserial ‘‘ archiptery- 
gial”” fin of Holoptychius. In Rhizodopsis and 
Glyptopomus they have given an account of the palate 
which considerably extends our knowledge of the 
structure of this region, and has a very important 
bearing on the problem of Amphibian ancestry. Fur- 
ther, they show that Ceratodus is derived by a process 
of specialisation and reduction from Dipterus, while 
Uronemus and Ctenodus are shown to belong to 
different lines of descent, and the conclusion is reached 
that Dipterus valencennes is the most primitive, as it 
is the oldest, known Dipnoan. 
Dr. C. A. Barper, sugar-cane expert to the Govern- 
ment of India, describes, in a recent number of the 
Memoirs of the Agricultural Department of India 
(Botanical Series, vol. viii., No. 3), some results of 
the work done at the Coimbatore Cane-Breeding Sta- 
tion since 1912, when, for the first time in India, 
seedling canes were raised successfully. The object 
of the work is to produce canes for cultivation in 
northern India under field conditions possible to the 
Indian peasant. The work is therefore limited in its 
NO. 2451, VOL. 98] 
“objective, and is also handicapped by unsuitability of 
soil, insufhcient chemical assistance, etc.; but in spite 
of these difficulties substantial progress has been made 
and certain provisional conclusions of general interest 
have been reached. Special attention may especially 
be directed’ to the various devices adopted to save 
labour and simplify the laborious business of selection. 
Arising out of this aspect of the subject, the possible 
correlation of morphological characters with richness 
of juice has been specially studied, and already Dr. 
Barber is able to say that “from this study it would 
appear that the seedlings in any general collection, 
with higher sucrose content, would be marked out’ by 
rather narrow, short leaves, but with a relatively high 
leaf module, with canes that might be thick or thin, 
but with a leaning towards the thin side, rather long, 
but not very, and with a~ moderately high cane 
module,”’. To this rule there are, however, many and 
notable exceptions, particularly in the case of definite 
crosses. It is clear from Dr. Barber’s memoir that 
the initial difficulties experienced at Coimbatore have 
been successfully overcome to a considerable extent, 
and the results obtained in the next few years should 
be of great value to the Indian sugar-cane industry, 
which has a great deal of leeway to make un before it 
reaches the level attained in other tropical sugar-pro- 
ducing countries, both on the scientific and the manu- 
facturing sides. 
In a paper read before the Franklin Institute in 
April Mr. C. J. Gadd, chief engineer of the American 
Iron and Steel Manufacturing Company, directed 
attention to the progress which has been made recently 
in the application of powdered coal to the heating of 
furnaces for metallurgical and other purposes. If 
slack coal is dried, pulverised until 83 per cent. of it 
will pass through a sieve with 200 meshes to the 
inch, and delivered to the furnace in this finely divided 
state, its combustion is completed while it is still in 
suspension, and as high a temperature is attained as 
with producer gas. From the experience gained with 
the apparatus used by the American company Mr. 
Gadd concludes that on further development powdered 
coal will entirely displace oil, tar, and producer gas 
in the fields in which they are now supreme. The 
paper is reproduced in the September number of the 
Journal of the- Franklin Institute. 
SomE definite evidence regarding the cause of the 
accident at the new Quebec Bridge on September 11 
has now reached this country, and is dealt with at 
considerable length by Engineering and the Engineer 
for October 13. The accident occurred while the cen- 
tral span was being raised into position, and the whole 
of this heavy portion of the structure fell into the 
river. Facilities were given to the representatives of 
the Engineering News to inspect the parts which failed, 
and from their report it appears that one of the four 
rocker bearings which supported the’ central span 
whilst in course of construction gave way when the 
span had been raised about 30 ft. Engineering hesi- 
tates to criticise, since the St. Lawrence Company had 
sought the advice of many able engineers, but at the 
same time considers that the design of the rocker 
bearings was very unsatisfactory. in view of the 
immense load of 1300 tons to be imposed on each. So 
soon as the span was lifted there was no longer any 
need for rocker bearings, as each corner of the great 
girder was suspended from the corresponding corner 
of one of the cantilevers. quite freely. It was only 
during transit on the barge that the rocker bearings 
could serve any purpose. Once the lifting was com- 
menced they were superfluous, and, as the event 
showed, an imminent source of danger. 
