72 
PACIFIC SCIENCE, Vol. XII, January, 1958 
natural magnetic spherules found in the sedi- 
ment. The yield was found to exceed 90 per 
cent. The results from the preliminary investi- 
gation, including a graph showing the size- 
distribution of the natural spherules, has been 
published by Laevastu together with Prof. O. 
Mellis of Stockholm (1955), who had given 
him valuable help in the microscopical study 
of the spherules. 
Our first objective when starting our work 
in November, 1954, was to perfect the method 
of extraction, involving also the preparation 
of the sediment samples obtained from the 
cores. The details of this preparatory work, 
carried out in the Oceanographic Institute 
with the excellent help of Mrs. Karin Romlin, 
technical assistant, will be given in a future 
publication by one of us. It will suffice to 
mention here that we found it necessary to 
pass an aqeous suspension of a quantity of 
the sediment, weighing from 200 to 700 grams 
(with the narrow corer the sample weight 
varied between 20 and 300 gr.), at least three 
times through the extractor. From the mag- 
netic particles thus obtained, which com- 
prised also a large proportion of nonspherical 
particles of terrestrial origin (largely mag- 
netite), those with strong magnetic proper- 
ties were separated out by means of a small 
electromagnet and then passed through sieves 
of different mesh, separating out the three 
size-classes: coarser than 60 /x, 60 /x to 30 /jl, 
and less than 30 /jl in diameter (see Figs. 1 and 
2). The products from this fractionation were 
then mounted between glass discs in a man- 
ner suitable for counting under the micro- 
scope. This latter operation was made by one 
of us in Stockholm, more recently assisted by 
fil. kand M. Nilsson. 
The operations required for a detailed ex- 
amination of the spherules, like polishing, 
X-ray examination, etc., were carried out in 
the Mineral ogical Institution of Stockholms 
Hogskola (Fredriksson, 1956). To its director, 
Professor S. Gavelin, we are much obliged 
for his support and advice. We are also in- 
debted to Professor F. Hecht of the II Chem- 
isettes Institut der Universitat Wien, for 
kindly carrying out for us micro-analyses for 
nickel and cobalt on some of the spherules 
submitted to him. His results proved nickel 
to be definitely present in a percentage vary- 
ing between 6 per cent and 15 per cent of the 
iron content. Recently Dr. A. Smales at Har- 
well kindly investigated samples of deep-sea 
spherules for nickel and cobalt by means of 
neutron-activation in the pile. 
In order to get comparable results from one 
sample to another, we found it advisable to 
concentrate the counts on black spherules of 
a diameter exceeding 30 /x. The spherules 
smaller than 30 fi are difficult to count and 
easy to overlook. Their contribution to the 
total weight of the spherules appears insig- 
nificant. Spherules of the greatest size, i.e., 
from 60 ^ to 250 ,u were also counted. Their 
contribution to the total number of spherules 
is not great, but, owing to their large size, 
their contribution to the total weight is con- 
siderable. 
The following table gives a survey of the 
cores from which we have extracted and 
counted spherules. 
COMMENTS 
core 71. This core is of special interest, 
as it was raised from the vicinity of Challenger 
Station 274 (S 07° 25' W 151° 15') where the 
depth was 2,750 fathoms or 5,030 m. The 
sediment in the surface has been characterized 
by Murray and Renard (1897) as Radiolarian 
Ooze with 3.89 per cent CaC0 3 . The core 
raised from the "Albatross” has a total length 
of nearly 10 m., from which sections of 26 to 
76 cm. in length were taken. 
The number of spherules per kg. of sedi- 
ment, free from salts and lime, varied from a 
maximum of 1,400 near the surface to a mini- 
mum of slightly more than 100 per kg. A 
second maximum of 600 spherules per kg. 
appears at the 10 metre level. Control counts 
made on halves of the section, cut lengthwise, 
gave fairly large variations in the proportion 
