3.5.4. The Deeply Probed Universe
Suggestions of very large scale structure (and by inference large
scale voids) has been seen in the recent Pencil Beam surveys
performed by David Koo and his collaborators. To date a total
of 5 very deep pencil beam fields have been done (Willmer et al. 1996)
at both the North and South galactic poles.
An initial analysis of about 400 galaxies (Broadhurst et al. 1990)
yielded very surprising results.
While the Slice surveys have shown a pattern of high and low
galaxy density areas on scales of tens of millions of light years
(or in redshift distance 1-5,000 km s-1), the deep pencil beam
surveys found separations between peaks in the galaxy distribution
of 10-20,000 km s-1. Moreover, these new surveys seem
to identify a scale of 
12,800
km s-1 that is present in the large scale galaxy
distribution. This scale is
comparable to the depth of the entire survey that discovered the
"Great Wall" of galaxies discussed above. To make matters worse,
the initial analysis show that the spacing between density peaks
was periodic. This kind of large scale "Cosmic Picket Fence" is
certainly at odds with the preferred model of a homogeneous and
isotropic universe. Critics (e.g., Ramella et al. 1992; Bahcall 1991)
suggested that the results were
a result of chance intersections of the pencil beam
with various rich groups or clusters of galaxies or is a rare manifestation
of a cellular distribution of galaxies (e.g., van de weygaert 1991).
To resolve this issue, Koo's international team gathered data for another 1000 galaxies in both directions. Their original results were not only confirmed, but they discovered that the peaks in the galaxy distribution extended away, from known rich clusters suggesting that the observed peaks were probably due to intersections with other "Great Walls" that permeate all space. With more data, the initial suggestion of periodic spacing between the peaks also vanished, to the collective relief of the cosmological community. That relief can be summed up in a prophetic phrase uttered by Marc Davis in 1990:
"If the galaxy distribution is truly periodic, it is safe to say we understand less than zero about the early Universe"
Knowing less than zero is an undesirable feature of any cosmological model yet, for a while, this is what the data suggested. The reconciliation between the Slice results and the deep pencil beam probes calls for a topology in the galaxy distribution that is "like a sponge". There are cells within larger cells, with the largest of the cells corresponding to a spacing of 10-20,000 km s-1 seen in the pencil beam data. Within this overall cellular topology, most of space is devoid of galaxies. The intersection between cells of hierarchical sizes seems to have produced some very large scale clustering. The evolution of this clustering pattern within the hierarchical cell sizes is a matter of much current research.