This method is briefly outlined in
Growing Gourmet
and Medicinal Mushrooms by Paul Stamets, but it is not illustrated in
the book. This method works very well for the isolation from bacterial
contaminants. Not only is the growth of bacteria inhibited by the
Antibiotic agar, the mycelium is forced to penetrate through the media
instead of across the surface, thus leaving the bacteria behind.
First you need the supplies:

0.3% peroxide solution -
optional (10ml 3% peroxide diluted with
90ml sterile water)
Sharp scalpel or knife
Inoculation loop
Fire (for flame sterilizing the loop)
2 Petris of Antibiotic Agar. (10mg Gentamicin per 200ml Agar and 2ml
3% peroxide is illustrated here – Peroxidated agar optional). Both
plates should be poured fairly thin to allow adequate room for the
sandwich.
First, flame your inoculation loop and cool it by dipping it into the
center of the transfer dish. Rescue the healthiest looking mycelium
from the wedge, as far away from the bacterial contamination as
possible. Try to get an area of 1cm square of mycelium, but do not get
any agar. Scrape the mycelium free using your inoculating loop. Here
scraping of mycelium from the surface of a plastic bag is illustrated.
Place the rescued mycelium in the center of the plate you used for
cooling the loop (into the melted divot) after dipping into the
peroxide solution. It is very important to not get any agar in the
previous step so the mycelium will lie flat, allowing it to be sealed
under the sandwich. Be very careful not to smear the transfer around on
the plate, or spread excess peroxide across the surface. You can omit
the peroxide but I use it as an additional safety measure. Be sure that
if you do use peroxide, you change it between each transfer to avoid
cross-contamination.
Cut a 1 inch (3cm) square from the other antibiotic agar plate you
have. It is best to pour the agar thin for this plate, so you can see
through it, and it is easy for the mycelium to grow through. I
recommend at least 1/16th inch minimum thickness (1.5-2mm) and a 1/8th
inch maximum (3-4mm). It is also important to make this agar not only
thin, but soft as well. You can accomplish this by reducing the agar
concentration, but I find the typical recipes to be fine for this
purpose (2% agar).
Place the square piece you removed earlier onto the rescued mycelium
fragment. Making sure to cover it all, and making sure all edges of the
agar sandwich are contacting the lower agar. Be very careful to not
smear anything around, spreading possible contamination. Basically you
want to seal the contaminated mycelium between the two pieces. Seal the
petri and incubate as required for your species.
After a few days mycelium will penetrate through the upper piece. It
may take as long as a week or more. Once it pokes through and appears
healthy, it is a good sign you now have pure mycelium (contaminant
free). Now it is important to keep a careful eye on the colonization to
watch for weak growth, or stalling.

(This is Hypsizygus ulmarius)
After a few more days, the mycelium will grow along the surface and
crawl down the edge to the base agar. It is important to watch for even
growth down onto the bottom layer. If it gets to that point and stops
(or stalls), it may indicate that bacteria has crawled out from between
the sandwich. Once it completely colonizes the dish, take a transfer
2-3cm away from the edge of the petri to ensure the cleanest tissue...
There you have it; it works very well to keep bacteria from crawling
along with your mycelium, and reduces the chances of airborne
cross-contamination when the bacteria are sealed under the agar. When
working with a contaminated petri, this method will guarantee you will
only need one transfer to antibiotic agar. Mycelium from this plate can
inoculate standard media recipes without antibiotics or peroxide.
By ATWAR