202 
a ! 
of the body capable of communicating the highest temperature, 
but by this, plus the minor radiators or convections of the cooler 
bodies? The words I have put in italics distinctly imply such 
an assumption. AG 
(© He seems to forget that I did, in the first place, observe and 
record the temperature of the surrounding medium, It was the 
19°C. which served as my starting-point. As no additional 
radiations were introduced beyond those of the flames to be 
experimented upon, and the blackened bulb of my thermometer 
was surrounded by polished reflecting metal surfaces on all sides, 
except that exposed to the flames, all the subsequent increments 
of heat were unquestionably due to the radiators from those 
flames, whether they came directly from the flames themselves 
or were received and reflected from the back and sides of the 
polished chamber, Fully admitting the desirability of a con- 
tinuous record of the heat thus communicated to the surroundings 
of the thermometer during the experiments, I nevertheless firmly 
maintain that, rude as it was, my apparatus (I refer, not to the 
thermometer, but to its adjuncts) was far superior to Capt. 
Ericsson’s. Mine was liable to a small source of error from a 
possible accidental irregularity of radiation by the thermometer 
bulb, but his was specially devised to ensure a large amount of 
such irregularity, continually increasing with the progress of the 
experiments, It is not a little surprising that so careful and 
luxurious an experimentalist as Capt. Ericsson should have 
overlooked the fact, that the very precautions which he so 
elaborately introduced to secure equal radiation from his bulb 
are precisely adapted to produce the contrary result. - 
The arrangements by which his thermometer is ‘‘ enclosed in 
an exterior vessel charged with water kept at a constant tempera- 
ture” of 60° by communication with a capacious cistern, directly 
violate the conditions demanded by the Newtonian law of radia- 
tion, of which Capt. Ericsson is so able a champion ; for as the 
experiment proceeds with an increasing number of flames, and 
consequent rising of the thermometer, this constant temperature 
of the water jacket goes on steadily augmenting the difference 
between the temperature of the bulb and that of its surroundings, 
and consequently secures just what it is intended to prevent, viz. 
a variable radiation. What is required to secure a constant 
degree of radiation from the bulb is not the constant tempera- 
ture of the surroundings, but a temperature steadily increasing 
at the same rate as that of the bulb, in order that the differential 
and not the adso/ute temperature of the surrounding medium, 
&c., should remain constant. ‘This was rudely obtained in my 
simple apparatus, as both the thermometer and its surroundings 
were simultaneously influenced by the same radiations. 
Capt. Ericsson takes great pains to controvert my ‘‘assump- 
tion that the intensity of a gas flame is proportional to the 
gas consumed.” This is unnecessary, inasmuch as I never 
made any such assumption, but have, on the contrary, endea- 
yvoured to prove that such cannot possibly be the case, by show- 
ing what becomes of the radiations from the interior of a large 
solid flame. If he will read chaps. 7 and 8 of ‘*‘ The Fuel of the 
Sun,” he will see how and why this has been done, and learn 
the true bearings of the experiments under discussion upon this 
subject. W, Martrieu WILLIAMS 
P.S.—The present is a suitable opportunity for asking a 
question which doubtless the philological readers of NATURE 
can easily answer. Many writers use the words ‘‘diather- 
mavcy,” ‘‘diathermavous,” ‘‘athermavous,” &c., rather than 
‘* diathermacy,” diathermous,” &c. Why is this? We do not 
say ‘‘thermaval” or “thermazometer,” &c. Why, then, 
should we depart from the analogy of ancient usage in con- 
structing the more modern compounds of the same root? 
Pollen~eaters 
Mr. Hart’s note in NATURE, vol. vii.p. 161, is interesting to 
those who have paid attention to the subject of fertilisation by 
insect agency, and would be still more so if he could furnish the 
names of the species of both plants and Syrf/ide that have come 
under his observation. 
May I take this opportunity of calling the attention of the 
readers of NATURE to a suggestion which I made some months 
since in the Journal of Botany, and which has at present met 
with no response? I believe no greater service could be ren- 
dered to this department of physiological botany than a series of 
observations on the species of insects which frequent and assist in 
the fertilisation of our wild flowers. I know of no such list even 
NATURE 
with respect to our commonest flowers. Here is a wide field for 
observation during the next season. 
London, Jan. 7 ‘ALFRED W. BENNETT 
P.S.—At the time of writing the above, I had not seen Dr, 
Buchanan White’s article in the January number of the ‘Journal 
of Botany,” on ‘‘The Influence of Insect-ageney in the Distribution 
of Plants,” an admirable introduction t@the series of papers I 
had in my mind. 
Welwitschia 
Ir you will kindly permit me, 1 wish to make an addition to 
your notice of my paper on ‘‘ Welwitschia,” read at the Linnean 
Society on the 19th ult. That paper was completed and put in 
Dr. Hooker’s hands about three months ago; and the reading 
of it was delayed until I had seen Strasburger’s recently pub- 
lished memoir on Coniferee and Gnetacex. After perusing that 
valuable work, I added a small appendix to my paper, and it is 
to the omission of the remarks contained therein that I wish to 
direct attention. 
In the description of the male flower, Strasburger and I al- 
most completely agree. It possesses two outer parts of the 
perianth, two inner parts, six stamens, which I believe to arise 
by branching from two primordial stamens, although Strasburger 
does not agree with me in this, and two ecarpels. The formula 
of the flower may be expressed thus :— 
Ca, Co, An,? Gn, 
In the female flower I had very great difficulty in coming to a 
conclusion as tothe value of the two outer parts, but she inner I con- 
cluded was a covering of the nucleus, an ovular integument, and 
not carpellary. There were only two ways of deciding what was 
the morphological significance of the two outer parts, either by 
comparison with the male’flower, or by comparing them with 
the parts in the flowers of Zphedra and Guetum. 1 applied to 
Dr. Hooker for specimens of these genera, and he has kindly 
promised to procure them for me. As Strasburger’s material for 
the examination of Ephedra and Gnetum was imperfect, it is 
still of importance to examine both in detail. Being, therefore, 
obliged to fall back on comparison with the male flower (the study 
of the development alone not being sufficient for the purpose), I 
described the two outer parts as forming a perianth, although I 
could not feel certain that I was correct in so doing, and could 
not explain the occurrence of the short stalk under them, no 
such stalk existing in the male. On looking at Strasburger’s 
figures of Ephedra, I at once saw that I had been in error in 
describing the outer parts as forming a perianth, and in the 
appendix stated that they were carfellary. 
The formula would therefore be :— 
Ca, Coy Any Gn, : 
The carpels, therefore, exist in both flowers ; but whereas in 
the male they are anterior and posterior, in the female they are 
lateral. Kindly make this correction, because I do not think 
that after Strasburger’s magnificent work, the Gymnospermous 
theory is for a moment tenable. 
Should any correspondent be able to obtain specimens of 
Ephedra and Gnetum for me, I would be greatly obliged, as I 
am desirous of completing my paper on ‘‘ Welwitschia” by a 
description of its embryogeny, as well as that of the other two 
genera. Specimens which have been put in absolute alcohol are 
by far the best for examination, but that, I fear, could only be 
obtained abroad with great difficulty. W,. R. McNas 
Dublin, Dec. 27, 1872 
Gauges for Ocean Rainfall 
In reply to Mr. Miller’s letter on ocean rainfall, in NATURE, 
vol. vii. p. 123, I beg to acquaint your correspondent that 
I have endeavoured to meet the difficulties he mentions, by de- 
signing two forms of rain-gauge for use on board ship. One 
is of a cylindrical form, and composed of a collector and re- 
ceiver, detachable from each other, and is suspended on gimbals 
in a frame or vexa. The rainfall may be estimated either by a 
glass scale at the sides, or by emptying the contents into a gradu- 
ated glass measure. 
A description of this instrument as above designed appeared 
in the Journal of the Scottish Meteorological Society for January 
1870, and was illustrated by diagrams. 
The other form consists of the cylinder asabove, divided into col- 
| ¥an, 16, I 873 
