274 CLINICAL BACTERIOLOGY AND H/EMATOLOGY 



It is usually over i, and, taking the average of a number of 

 cases, it increases in proportion to the amount of reduction of 

 the corpuscles, thus : 



When you find a case with a high colour-index, it should 

 immediately raise a suspicion of pernicious anaemia. Turn 

 again to the preparation in which you have counted the reds 

 in the Thoma-Zeiss haemocytometer, and look for unusually 

 large corpuscles (megalocytes), which, according to Ewing, 

 should form 35 per cent, of all corpuscles, or the diagnosis is 

 to be made with caution. Then count the leucocytes in the 

 same preparation. Leucopenia is very characteristic; if the 

 number exceeds 6,000, pernicious anaemia is unlikely, unless 

 inflammatory complications are present. Then make a differ- 

 ential count on a stained film, looking out for me galo blasts 

 and normo blasts as you do so; in pernicious anaemia there is 

 almost always a relative lymphocytosis,'and. the diagnosis is 

 unlikely with the lymphocytes much below 40 per cent. If 

 you have not yet seen a megaloblast, continue to search for 

 them, as they are usually present in pernicious anaemia of 

 moderate severity, and comparatively rare in other conditions, 

 except in children. The significance of the discovery of a 

 single megaloblast will depend on the other findings; if these 

 point to pernicious anaemia, the megaloblast may be taken as 

 clinching the diagnosis, but if they are not of this nature its 

 importance is much less. Do not exclude pernicious anaemia 

 because no megaloblasts are found. According to Ehrlich 

 and others, they always exceed the normoblasts in numbers, 

 but this is not a safe guide, as in some cases you may find 

 normoblasts alone on some occasions, and megaloblasts and 

 normoblasts in differing proportions on others. Polychro- 

 masia and granular degeneration and poikilocytosis are com- 



