PRACTICAL EXERCISES 61 



sulphate. Filter. Precipitate the serum-albumin from the filtrate by 

 saturating with ammonium sulphate crystals. 



(5) Dilute serum with ten to twenty times its volume of distilled 

 water, and pass through it a stream of carbon dioxide. The serum- 

 globulin is partially precipitated. This is . the starting-point of a 

 method said to be the best for obtaining pure serum-globulin. 



(6) Acidulate some serum with dilute acetic acid and boil. Filter 

 off the coagulum, and to the filtrate add silver nitrate. A non- 

 proteid precipitate insoluble in nitric acid but soluble in ammonia 

 indicates the presence of chlorides. 



6. Enumeration of the Blood-corpuscles. Use the Thoma-Zeiss 

 apparatus (Fig. 13). Prick the finger to obtain a drop of blood. 

 Suck the blood up into the capillary tube S to the mark i.* Wipe 

 off any blood which may adhere to the end of the tube. Then fill 

 it with Hayem's solution (p. 29) or 3 per cent, sodium chloride to 

 the mark 101. This represents a dilution of 100 times. Mix the 

 blood and solution thoroughly, then blow out a drop or two of the 

 liquid to remove all the solution which remains in the capillary tube. 

 Now fill the shallow cell B with the blood mixture. Slide the cover- 

 glass on, taking care that it does not float on the liquid, but that the 

 cell is exactly filled. Put the slide under the microscope (say Leitz's 

 oc. III., obj. 5), and count the number of red corpuscles in not less 

 than ten to twenty squares. The greater the number of squares 

 counted, the nearer will be the approximation to the truth. Now 

 take the average number in a square. The depth of the cell is 

 to- mm., the area of each square ?&& sq. mm. The volume of the 

 column of liquid standing upon a square is ^fou cub. mm. One 

 cub. mm. of the diluted blood would therefore contain 4,000 times 

 as many corpuscles as one square. But the blood has been diluted 

 100 times, therefore i cub. mm. of the undiluted blood would con- 

 tain 400,000 times the number of corpuscles in one square. Suppose 

 the average for a square is found to be 13. This would correspond 

 to 5,200,000 corpuscles in i cub. mm. of blood. 



7. Opacity of Blood. Smear a little fresh blood on a glass slide, 

 and lay the slide on some printed matter. It will not be possible to 

 read it, because the light is reflected from the corpuscles in all 

 directions, and little of it passes through. 



8. Laking of Blood. (i) Put a little fresh blood in three test- 

 tubes, A B and C. Dilute A with an equal volume, B with two 

 volumes, and C with three volumes, of distilled water, and repeat 

 experiment 7. The print can now be read probably through a 

 layer of A, but certainly through B and C, since the haemoglobin is 

 dissolved out of the corpuscles by the water and goes into solution, 

 the blood becoming transparent or laked. That the difference is 

 not due merely to dilution can be shown by putting an equal quantity 

 of blood in two test-tubes, and gradually diluting one with distilled 

 water and the other with a 0^9 per cent, solution of sodium chloride, 

 which does not dissolve out the haemoglobin. Print can be read 



* If the tube has not been properly filled, blow the blood out immediately. 

 On no account permit it to clot in the capillary tube. 



