This paper introduces type classes as a solution to ad-hoc polymorphism, extending the Hindley/Milner type system to support overloading. Type classes allow functions to be defined for multiple types, with different implementations for each type, and generalize the concept of "eqtype variables" from Standard ML. They provide a way to handle issues in object-oriented programming, bounded type quantification, and abstract data types. The paper presents an informal introduction to type classes and defines them formally using type inference rules. Type classes are motivated by the limitations of ad-hoc polymorphism in languages like Standard ML and Miranda, where overloading of arithmetic and equality operations leads to complex and inefficient solutions. Type classes provide a uniform mechanism for handling these issues by grouping related operators and allowing translation to a form that does not use overloading. This translation is done using a method that passes method dictionaries at runtime, enabling efficient and type-safe handling of overloaded functions. The paper includes examples of defining arithmetic operations and equality using type classes, demonstrating how they can be applied to abstract data types. It also discusses the translation of these examples into a form that can be processed by the Hindley/Milner type system. The paper further explores the use of subclasses, where a class can inherit from another, and discusses the relationship between type classes and other concepts like bounded quantification and subtyping. The paper concludes that type classes offer a flexible and efficient approach to ad-hoc polymorphism, with potential applications beyond Haskell, such as in Standard ML. It also highlights the importance of further research into the relationship between type classes and other programming concepts, as well as the trade-offs between different approaches to polymorphism.This paper introduces type classes as a solution to ad-hoc polymorphism, extending the Hindley/Milner type system to support overloading. Type classes allow functions to be defined for multiple types, with different implementations for each type, and generalize the concept of "eqtype variables" from Standard ML. They provide a way to handle issues in object-oriented programming, bounded type quantification, and abstract data types. The paper presents an informal introduction to type classes and defines them formally using type inference rules. Type classes are motivated by the limitations of ad-hoc polymorphism in languages like Standard ML and Miranda, where overloading of arithmetic and equality operations leads to complex and inefficient solutions. Type classes provide a uniform mechanism for handling these issues by grouping related operators and allowing translation to a form that does not use overloading. This translation is done using a method that passes method dictionaries at runtime, enabling efficient and type-safe handling of overloaded functions. The paper includes examples of defining arithmetic operations and equality using type classes, demonstrating how they can be applied to abstract data types. It also discusses the translation of these examples into a form that can be processed by the Hindley/Milner type system. The paper further explores the use of subclasses, where a class can inherit from another, and discusses the relationship between type classes and other concepts like bounded quantification and subtyping. The paper concludes that type classes offer a flexible and efficient approach to ad-hoc polymorphism, with potential applications beyond Haskell, such as in Standard ML. It also highlights the importance of further research into the relationship between type classes and other programming concepts, as well as the trade-offs between different approaches to polymorphism.