a 
ir] BIOLOGY OF THE RHINE SALMON. 437 
the rest, its ovaria will certainly be found to be below the average size, 
and its eggs will in some cases be small, probably retarded in their 
growth. 
If, moreover, it could be shown that these salmon, which have reached 
their sexual maturity in the sea, constituted an anyway considerable frac- 
tion of all the salmon which ascend from the sea, it ought to be impossible 
to prove that the summer salmon, up to the spawning season, lose as much flesh 
as they gain in ovarium substance; there certainly ought to be a deficit on 
the side of the summer salmon. 
On account of the extraordinary scientific importance attaching to a 
certain proof of such a change of substance in the animal kingdom, I 
have for two and a half years (since November, 1877), as long as the 
salmon fisheries lasted, with hardly any interruptions, daily weighed 
and measured salmon, and portions of salmon, on the most extensive 
scale; the rich material on which I could operate being very kindly 
furnished by Mr. Glaser. In weighing I used altogether an admirably 
constructed pair of English spring scales, which through many years 
showed no sign of variance, and which in 10 kilograms distinetly indi- 
cated a difference of 20 grams. The measurements were made with a 
very simple apparatus, consisting of a board with a scale marked upon 
it, and another piece of board attached to it at a right angle, which 
could be moved up and down until it reached the back point of measur- 
ing (now recognized as the most reliable)—the point where the body 
tapers off and where it again begins to spread a little to form the caudal 
fin. From September, 1878, I also invariably determined the length of 
the nose (respectively, its horizontal projection, when the head is fixed 
in a horizontal position). In the male fish I considered as the “length 
of the fish” only the distance from the nostril to the root of the tail. 
The observations were taken very carefully every morning by Mr. Jacob 
Weidmann, the assistant of our Physiological Institute, generally in 
Mr. Glaser’s presence, and were tested by frequent observations of my 
own, which were entirely satisfactory. No error of any consequence 
was discovered among a hundred figures thus examined; and if ever 
statistical data deserve to be called reliable, this term must certainly 
apply to the data obtained in the manner described. 
T’rom among the female fish of one and the same year (1878), begin- 
ning with May or June, when the sex could be distinctly recognized, I 
selected specimens as nearly as possible of equal length, the greatest 
difference of length not amounting to more than 10 to 12 millimeters. 
The total number was divided into 2 to 4 groups, arranged in chrono- 
logical order, and for every group the average date was calculated. The 
comparison of the averages of the groups furnished the average change 
of weight of the fish in the periods included in the average dates. The 
same operation was gone through with six groups of as many different 
lengths, comprising in all 470 fish, all belonging to the same year (1878), 
and all caught between Basel and Laufenburg. From the changes of 
