Cluster Sampling

Relies on natural groups
The items are chosen in clusters, the ones close to each other
If necessary the clusters can be further divided into smaller clusters
For example you could select any 6 sequential items
Within each cluster items are then chosen by simple random sampling or some other method.

Advantages

1) Easier and quicker to identify the items

Disadvantages

1) Items close to each other may be very sumiliar and less likely to represent the whole population.
2) Larger sampling error than simple random sampling