VII.] ESTIMATION OF HAEMOGLOBIN. 65 



If 130 Ccytorneter) corresponds to 190 (cliromometer) then no cytometer 

 (?.e,, graduation corresponding to 100 parts of hasmoglobiu) corresponds to it 

 chromometer graduation : 



190. no 20,900 ,^ ^ 



no : 190 = 1 10 : a; . . a; = —2 = — 'y — = 100.7. 



•" 130 130 



Blood containing 100 ])arts luTemoglobin will correspond to 160 of the chromo- 

 meter scale, and beginning with this number as a basis, with the aid of our 

 formula it is easy to cojistruct a table showing the relation. 



Whilst the value of the cytometer scale remains the same for every instru- 

 ment, the chromometer scale varies with each instrument, as the colour- 

 intensity of the glass is not necessarily the same in all. But it is easy to 

 construct a scale for each instrument by investigating a specimen of blood 

 and comparing it with the cytometer graduation as indicated in the foregoing 

 paragraph. 



Frecautiaiis to he Observed in Using the Instrument. — The exact quantity of 

 the several fluids must be carefully measured ; evaporation must be prevented 

 by covering the blood-mixture. Further, do not look at the fluid too long at 

 a time, as the eye becomes rapidly fatigued. Further, the operation must be 

 carried out not too slowly, as the saline solution only retards the coagulation 

 of the blood, and does not arrest it. 



In cases of leukaemia, where there is a large number of white corjmscles 

 rendering the mixed fluid opaque, the corpuscles may be made to disappear by 

 adding a drop of a very dilute caustic potash. If the opacity does not disappear 

 by the addition of this substance, then the o])acity is due to the presence of 

 fatty granules in the blood, so that by this means we can distinguish lipttmia 

 from leukaemia. 



Bizzozero claims that when the instrument is used as a cytometer the mean 

 error is not greater than o. 3 per cent 



10. Preparation of Hsemoglobin {clog^s or Iwrse's blood). — Centrifugalise 

 filtered fresh defibrinated dog's blood, and when the corpuscles have subsided 

 pour off" the clear serum. Mix the corpuscles with .5-2 per cent, solution of 

 NaCl, and centrifugalise again. Repeat the process until the washings con- 

 tain only a trace of proteid, or begin to be tinged red from the solution of the 

 blood-corpuscles. 



Mix the magma of corpuscles with 2-3 volumes of water saturated with 

 acid-free ether. The corpuscles swell up, become almost invisible, and the 

 solution becomes clear. "With the utmost care add, stirring all the time, i per 

 cent, solution of acid sodic sulphate until the blood appears turbid like fresh 

 blood. The stromata of the corpuscles are thereby caused to shrivel, and when 

 they are centrifugalised for a long time, they run together, and can thus be 

 separated. Pour ofl* the clear fluid, cool it to 0°, add one-fourth of its volume 

 of pure alcohol previously cooled to o" or lower. Shake up the whole, and let 

 it stand for twenty-four hours at 5°-i5''. As a rule, the whole passes into a 

 glittering crystalline mass. Place it in a filter cooled to 0°, and wash it with 

 ice-cold 25 per cent, alcohol. Redissolve the crystals in a small quantity 

 of water, and recrystallise with alcohol as before. The crystals are spread on 

 plates of porous porcelain, and dried in a vacuum over sulphuric acid. 



11. Amount of Hsemoglobin in Blood— Colorimetric Method (Hoppe- 

 Seyler's method). — A standard solution of pure haemoglobin diluted to a 

 known strength is used, and with this is compared the tint of the blood 

 diluted with a known volume of distilled water. 



(a.) A standard solution of hsemoglobin of known strength is supplied (sMj^ra), 



B 



