heparin. The data in table 4 show as in table 3 

 that pretreated blood gave lower hematocrits . 

 It is interesting to note that values were some- 

 what lower in 32 mm capillaries. 



In another series blood from 8 trout of 

 each species was pretreated with heparin in test 

 plates and at the same time blood from the same 

 trout was introduced directly to plain capillaries. 

 After the untreated blood had coagulated in plain 

 capillaries, all capillaries were centrifuged 

 simultaneously and the columns of packed red 

 cells and the columns of packed clot were meas- 

 ured. The results are presented in table 5. They 

 should be compared with those presented in tables 

 3 and 4. It is evident from this comparison that 

 the commercially available heparinized 75 mm x 

 1.1 - 1.2 capillaries are not entirely satisfactory 

 for exact determination of hematocrits in trout, 

 giving only slightly better results than plain capil- 

 laries. Comparison of these microhematocrits 

 with data shown in table 3, shows that the com- 

 mercial heparinized capillaries are entirely 

 satisfactory with human blood for which they are 

 prepared. 



Hematocrit values with trout blood obtained 

 in commercial heparinized capillaries were found 

 to be 7 to 18 percent higjier than readings obtained 

 with pretreated blood in plain capillaries. There- 

 fore until this difficulty is removed, the hemato- 

 crit values obtained with 'the 75 mm 1.1 to 1.2 

 commercial heparinized capillaries should be 

 corrected accordingly. 



Results from determination of hemoglobin 

 in trout blood and oxyhemoglobin are presented in 

 table 6 . Data may be of value to persons who are 

 using these two methods for fish blood examina- 

 tion. 



DISCUSSION 



It has been found with trout, as with hu- 

 mane, that the hematocrit is a simple, accurate 

 and rapid method for detection of ordinary anemia . 

 One must keep in mind, however, that there are 

 many types of anemia (Wells, 1956; Wintrobe, 1958) 

 and it is likely that in an examination of fish blood, 

 determination of hematocrit cannot always be used 

 as the only hematological method. Microhemato- 

 crit is as accurate as macrohematocrit (McGovern 



et a]_. , 195^ it requires only one or two drops 

 of blood and the ultra micro method of Strumia 

 et al. (1954) can be run with even less blood. 

 It is evident from the excellent comparative 

 hematological examination reported by Wintrobe 

 (1934) that hematocrit is a relatively uniform 

 value in all vertebrates and that it correlates 

 very well with red cell counts , hemoglobin and 

 other hematological values. Wintrobe found 

 that differences in hematocrits were very small 

 in comparison to great differences in number and 

 size of red cells. Also the differences in mean 

 corpuscular hemoglobin corresponded inversely 

 to differences in number of red cells . Therefore 

 the amount of hemoglobin per unit volume of blood 

 varied slightly. The mean corpuscular hemoglobin 

 concentration was the most uniform constant in all 

 vertebrates examined. There is a very close re- 

 lationship between the amount of hemoglobin and 

 hematocrit per unit of blood volume (Wintrobe, 

 1934). There is little doubt that the microhemato- 

 crit is a useful procedure in routine and special 

 examination of fish blood. 



In order to use hematocrit for evaluation 

 of condition of fish in hatchery production, manage- 

 ment, or research it is necessary to know the 

 "normal" hematocrit values. Most of the humans 

 and land vertebrates live in an atmosphere of 

 uniform composition . The gill breathing animals 

 may be exposed in nature to an extreme variety of 

 environmental conditions, especially availability 

 of oxygen and concentration of carbon dioxide . 

 Therefore one must expect that under different en- 

 vironmental conditions there will be considerable 

 hematological variation in otherwise "normal" 

 animals. While it should be possible to establish 

 the "normal" hematocrit values for different 

 species of fishes, the amplitude of normal values 

 will have to be wider than that for terrestrial and 

 particularly for warm blooded animals . 



Amount of dissolved oxygen (Hall et al. , 

 1926; Phillips, 1947; Dombrowski, 1953); season 

 of the year (Ivlev, 1957; Yokoyama et al., 1947), 

 physiological activity (Dubravko, 1956), temper- 

 ature (Scholander, 1957), pollution (McCay, 1929) 

 and perhaps other factors have a pronounced in- 

 fluence on the morphological and chemical 

 composition of fish blood. 



Because even in "normal" fish the composi- 

 tion of blood may show considerable variation, it 



