curve for seawater and the close physical proximity of the two 

 chemistries. 



When we compare the results of our work at Axial Volcano with 

 recent experimental results on the chemical effects of phase 

 separation of hydrothermal fluids (Bischoff and Rosenbauer, 1987), 

 there are some differences, but the major trends are the same. 

 The pseudo-conservative element chloride is somewhat higher in the 

 low-salinity Axial endmember than found in the experimental vapor 

 phase, while the Axial volatile components are considerably 

 elevated over the levels in the experimental vapor phase. Some of 

 the factors which contribute to the differences between the 

 experimental results and our field results are: 1) different 

 pressure and temperature conditions, 2) incomplete segregation of 

 vapor phase from parent fluid in the field, 3) differences in 

 chemistry of the basalt and evolved hydrothermal fluid between the 

 experiment and the field, and 4) closed experimental system 

 versus possible open-system boiling in the field. 



It is anticipated that the complete analysis of the much 

 larger 1987 vent fluid data set will help to align the above 

 observations with a single, self-consistent model for the Axial 

 hydrothermal system. 



Submersible-coupled In situ Sensing and Sampling System (SIS 3 ) 



Two rather obvious weaknesses of the 1986 Axial vent fluid 

 data set are the lack of relatively undiluted (i.e., >^ 90% 

 hydrothermal, Mg _< 5 mmol kg -1 ) vent fluid samples and reliable 

 temperature data, especially temperatures measured coincident with 

 sample collection. Near-pure hydrothermal fluid data substan- 

 tially enhances our ability to extrapolate to endmember values 

 and, in fact, may serve as reasonable direct measures of these. 

 For non-conservatively behaving plume species, like Fe and Ba, 

 undiluted vent fluid data are essential to a nonambiguous end- 

 member determination. The vent fluid data set assembled by Karen 

 Von Damm (1983) for the East Pacific Rise (EPR), 21°N hydrothermal 

 fields is one of the most complete yet reported. The Fe versus Mg 

 data for this site are shown in Figure 5. By the aforementioned 

 criteria, only 27% (n=20) of the vent fluid samples were suitable 

 for use in the determination of the endmembers for Fe (and, 

 similarly, Mn, Co, Ni, Cu, Zn, Ag, Cd, Pb, Hg, H 2 S, As, Se, pH, 

 and Alkalinity). It is likely that some of the more diluted 

 samples shown in Figure 5 are artifacts of sampling: i.e., the 

 sampler snorkel tip was not in the position of highest 

 temperature, hence, maximum fluid purity, during piston 

 retraction. The sensing of temperature during sampling would 

 increase the efficiency with which undiluted fluids are recovered. 



44 



