p o s s i b i l i t y , the Single Stage Model, in the context o f which the element was suggested as the l i g h t element in the core (Ringwood, 1958, 1959, 1960,
19661,b, 1973), faces several d i f f i c u l t i e s which cannot e a s ily be overcome (Ringwood, 1977b, 1979).
The Revised Homogeneous Accretion Hypothesis can s a t i s f a c t o r i l y explain the "overabundance" o f s id e ro p h ile elements, and the high Fe3+/Fe2+ r a t i o in the upper mantle, p re c is e ly because i t involves in tro d u c tio n o f oxygen as the p r in c ip a l l i g h t element in the core.
6.2 HOMOGENIZATION PROCESSES In tro d u c tio n
I f homogeneity w ith respect to compatible elements is in fe rre d f o r the mantle, then there are several possible in te r p r e ta t io n s (a) i t may be
a primary fea ture o f the mantle; (b) the cause may be s o lid - s t a t e
d i f f u s i o n ; or (c) l i q u i d - s t a t e d i f f u s io n ; and (d) convection may be
responsible (Hutchison e t a l . , 1975; Hofmann and H art, 1975, 1978; Hof
mann and Magaritz, 1978; Ringwood, 1979). I t is necessary to show th a t
the scale o f observation o f degree o f heterogeneity l i e s outside the scales o f operation of remixing processes, before i t can be concluded th a t chemical u n ifo rm ity is an o r ig in a l fea ture o f the mantle (Hofmann and Magaritz, 1978).
D iffu s io n
The region o f the upper mantle where S wave v e lo c it y , and perhaps also P wave v e lo c it y pass through a minimum may be a region containing a
small degree o f p a r t ia l melt. The region is known as the low v e lo c it y
zone ( L . V. Z . ) , and f o r i t to be p e rs is te n t over geological timescales, w ith a homogeneous d i s t r ib u t i o n o f m e lt, a melt f r a c t io n only on the order
o f 1.0-0.1% i s allowed (Walker et a l , 1978). This can s t i l l explain the
seismic pro pe rties of the zone (Anderson and Spetzler, 1970). The
assumption behind t h i s i s , however, th a t the melt occurs as f i l m s , or in cracks, but Waff and Bulau (1977) propose t h a t the geometry is th a t of tubules on 3-grain contacts, which may require at l i t t l e more melt
(Walker e t al , 1978). However, d i f f u s i o n is u n l i k e l y to be s i g n i f i c a n t on
a large scale i f the melt does not provide a s h o r t - c i r c u i t around grains. I t has been suggested th a t on a regional (km) scale, d i f f u s i o n in conjunction with convection could homogenize the mantle (O'nions and
Pankhurst, 1974). But assuming the d i f f u s i v i t i e s o f d i v a le nt cations in
p e r i d o t i t e , composed dominantly of o l i v i n e , are of the order o f D = 10“ 13 to 10“ 14 cm2/ s e c . , (Hofmann and Hart, 1978), T = 1000°C-1200°C, and using the r e l a t i o n x = (Dt)*5-where x is the tr an spo rt distance, t is time, and D the d i f f u s i o n c o e f f i c i e n t , the c h a r a c t e r i s t i c distance at which s o l i d -
state d i f f u s i o n could occur is 10-15cms in 10s-109 y rs . Clark and Long
(1971) found a value o f D = 10"14cm2/sec f o r Ni2+ d i f f u s i o n in o l i v i n e at 1149°C.
Thus, local d is e q u i l i b r iu m could e x i s t in the mantle s i l i c a t e s with respect to d i v a l e n t catio ns, and be maintained to the present day in a s o l i d mantle (Hofmann and Hart, 1978).
When a b a s a l t i c melt is present, a short c i r c u i t is provided f o r d i f f u s i n g cat io ns , which applies even when the melt is only a thi n f i l m ,
wetting the grain boundaries. I f a d i f f u s i o n c o e f f i c i e n t o f D = 10“ 6 - 10"7
cm2/sec is assumed, which is typ ic a l of d i f f u s i o n in dry b a s a lt ic melts, a volume o f mantle o f the order o f 1km would be homogenzied in 109 years
(Hofmann and Hart, 1975; Magaritz and Hofmann, 1977).
Watson (1979) found d i f f u s i v i t i e s f o r Cs in wet g r a n i t i c melts 3 to 4 orders of magnitude higher than f o r dry systems, r e s u l t i n g in 30-100