need much more information before anyone can begin to help :)
do you mean finding 400 records from your overall set of records - what criteria do you want to 'find' ?
or do you mean finding 300 unique records in a set of 400 - again what criteria makes a record unique? ID number? name?
how are you adding records? If your table has an ID field that is a serial number which updates automatically upon creation of a new record, and only allows unique values, then yes a new record will be unique by virtue of its serial number
If you want to remove the 100 duplicate records permanently from your table, there are several methods that can do that. Here's one method:
Save a copy of your database file and delete all records from this table in the copy. If there is only one table in your database, use the "clone" option to both save a copy and get it with no records in this table.
In the copy, use field options to specify "unique values" "validate always" on the field that contains the duplicate values in your table in the original copy. If there are several fields that make up a combined value that must be unique, add a new text field and give it an auto- enter calculation that combines the values of all of these fields and specify these field validation options on this new field.
Then use Import Records|File to import your data from the original file into this table of the new copy. Be sure to select the matching names option to map each field to it's matching field of the same name and be sure to enable "auto enter options" in the small dialog that pops up just before the import starts.
With these settings, the 100 records with duplicate values will be not be imported during the import and you will get a table with only unique records.