I 
Annual Report for 1887 of the Consulting Chemist. 299 j 
determined when in seed, as in the red clover (Trifolium pra- j 
tense) there is a single seed in the pear-shaped pod, and the pod i 
opens by the top falling off, while in cowg-rass (Trifolium ■ 
medium) there are two seeds in the oblong pod, and the pod opens ^ 
along the edges like the pod of a pea. The seed also is more 
heart-shaped than in the red clover. Sinclair recognised the 
importance of cowgrass sixty years ago, but he found it hard 
then to purchase the seed — now it is impossible. All the seeds 
now sold under the name of cowgrass belong to varieties of red 
clover, and these differ among themselves just as the varieties 
c# cultivated wheat differ. I have reason to hope that the true \ 
cowgTass may be soon available to the farmer. j 
Different estimates are formed of the germinating value of > 
mangold seeds. They are obtained from the seed-merchant, | 
not as separate seeds like clover or turnip, but enclosed in a 
compound fruit, each containing two or three seeds covered by 
the enlarged fleshy base of the calyx of so many flowers. Each i 
fruit should consequently produce two or three plants. These ' 
plants when they grow are, however, j^roduced so close to- 
gether that they are of no practical use to the cultivator, and 
have to be carefully thinned out. In testing the mangold seed 
I have determined whether each compound fruit would produce 
a plant ; others have taken each seed that germinated into ac- " 
count, and consequently record good mangold ''seed" as grow- i 
ing — say, 180 per cent. This, of course, means that of the seeds I 
in 100 compound fi'uits, amounting in all to 200 or 250, 180 
have grown. As the farmer must, however, consider each com- 
pound fruit as if it were a single seed, it seems the only fair 
way to test it from that point of view, and so to determine how 
far each " seed " supplied by the merchant is available to pro- 
duce a i^lant. 
XVII. — Annual Report for 1887 of the Consulting Chemist. 
By Dr. J. Augustus Voelcker, B.A., B.Sc, 12 Hanover 
Square, W. 
The number of samples analysed in the Society's laboratory 
during the past year was, in all, 1,615; of these 1,536 were 
analyses made for members of the Society, 56 were in connec- 
tion with the Woburn experiments, and 23 analyses on behalf 
of local Agricultural Societies carrying out experiments in 
conjunction with this Society. At the end of this Report is 
given a list of the diffei'ent kinds of analyses, and their number. 
Of the 1,615 samples analysed, nearly 700 have been of 
feeding-cakes. Bj' far the larger proportion consisted of linseed- 
