78 Fred C. Rohde and Rudolf G. Arndt 



body weight (W) is: Y = 49.19 + 31.43 W. The correlation coefficient of 

 0.77 is significant (P< 0.001), and indicates a positive relationship 

 between ova number and gravid female weight. 



Mature ova (large and orange) were not observed in 235 females 

 until January. In 1974, mean diameter of ova from 14 preserved females 

 increased from 1.18 mm in January to 2.07 mm just prior to spawning in 

 March. The largest ovum measured 2.4 mm (early March). Values in 

 1975 (N - 23 females) fell within these ranges. In March 1980 (N = 20 

 females) the range was 1.8 to 2.5 mm (x = 2.15 mm). There was no 

 difference in mean ovum diameter between ovaries (N = 22 females). 

 Ludwig and Norden (1969) gave ova diameters of 1.50 to 2.06 mm (x = 

 1.88 mm) in Wisconsin, and Hann (1927) of 2 to 2.5 mm in Michigan. 

 Nagel (1980) observed a larger average ovum size (3.32 mm) in Tennessee 

 C. bairdi and suggested that it correlated with the smaller number of ova 

 produced. 



The relationship between diameter of mature ova and SL is signifi- 

 cant (r = 0.55; P<0.01). This is in agreement with Hoar (1957) who 

 stated that mature ova size depends both on size of parent and on nutri- 

 tion during the pre-spawning period. 



In seven females, the right ovary was usually longer (x = 9.5 mm) and 

 narrower (x - 4.2 mm) than the left (x = 8.7 and 5.0 mm, respectively). 

 The right contained a mean of 41.2 ova, the left 39.9. There is no signifi- 

 cant chi-square difference from the expected 1:1 ratio (P>0.05). 



A total of 2 egg clusters was found in 1980, and 37 in 1981. The 

 larger number was due to additional egg attachment sites provided in 

 February 1981, and to more extensive searching. In 1980, the first cluster 

 was found on 15 March and the second on 22 March (deposited in the 

 interim). In 198 1 , the first seven clusters were found on 27 February, five 

 others on 7 March, and another six on 14 March. Additional clusters 

 found on subsequent dates resulted (at least primarily) from searching in 

 areas not examined earlier, rather than from later egg deposition. All 

 clusters were found in and near two gravel riffles in the sampling area 

 (Fig. 1), and to about 30 m upstream. All were in areas of fast current, 

 usually near the stream center, and over gravel or gravel-sand substrate. 

 Water depth by the eggs (on several dates in early 1980 and 1981) ranged 

 from 3 to 50 cm (x = 20.0 cm, N = 40 measurements). 



With one exception, every cluster was attached to the underside of a 

 submerged object. The exception was in a hollow on the upper surface of 

 a submerged log. One cluster was under a beer can and another under a 

 bottle, and three others were under waterlogged wood: a branch, a piece 

 of board, and a piece of cut log. Favored egg-attachment sites were pieces 

 of concrete block and bricks, and 23 (58.9%) of the clusters were found 

 on such objects. Ten clusters (25.6%) were on ironstone and silicified 



