There is a subtle difference between weakly (strongly) typed [programming language|programming languages] and dynamically (statically) typed languages.

The strong/weak classification refers to [type checking], and whether [type errors] are always found. If all type errors are found (either at [run time] or [compile time]), then the language is [strongly typed]. If few type errors are checked, then the language is considered to be weakly typed.

In this sense, no language is truly strongly typed, even [Ada] allows type checking to be suspended in certain cases. [Pascal] and [Java] are nearly strongly typed, while [C] and [C++] are weakly typed because they allow passing [parameters] of unknown types to functions.

The static/dynamic axis refers to when a [variable] is [type binding|bound] to its type. [Statically typed] languages require variables to be bound to their types at compile time, thus there is often need to declare the type of the variable in the code. [Dynamically typed] languages, on the other hand, bind the variables to their types only when a value is [assigned] to a variable, thus allowing a variable to hold values of different types at different points in a programs execution.

[Java], [C], [C++], and even [Perl] (to a certain extent) are statically bound languages, while [APL], [SNOBOL], and [Ruby] are dynamically bound languages.

Reference:
Robert W. Sebesta, Concepts of Programming Languages, 4th Ed.