1 Scope
2 Normative references
3 Subprograms provided
4 Instantiations
5 Implementations
6 Exceptions
7 Arguments outside the range of safe numbers
8 Method of specification of functions
9 Branch cut and domain definitions
10 Range definitions
11 Accuracy requirements
12 Overflow
13 Underflow
14 Specifications of the functions
    14.1 SQRT Square root
    14.2 LOG Natural logarithm
    14.3 EXP Exponential function, complex argument
    14.4 EXP Exponential function, imaginary argument
    14.5 "**" Exponentiation operator, complex to
                  power arguments
    14.6 "**" Exponentiation operator, complex to real
                  power arguments
    14.7 "**" Exponentiation operator, real to complex
                  power arguments
    14.8 SIN Trigonometric sine function
    14.9 COS Trigonometric cosine function
    14.10 TAN Trigonometric tangent function
    14.11 COT Trigonometric cotangent function
    14.12 ARCSIN Inverse trigonometric sine function
    14.13 ARCCOS Inverse trigonometric cosine function
    14.14 ARCTAN Inverse trigonometric tangent function
    14.15 ARCCOT Inverse trigonometric cotangent function
    14.16 SINH Hyperbolic sine function
    14.17 COSH Hyperbolic cosine function
    14.18 TANH Hyperbolic tangent function
    14.19 COTH Hyperbolic cotangent function
    14.20 ARCSINH Inverse hyperbolic sine function
    14.21 ARCCOSH Inverse hyperbolic cosine function
    14.22 ARCTANH Inverse hyperbolic tangent function
    14.23 ARCCOTH Inverse hyperbolic cotangent function
Annexes
A Ada specification for GENERIC_COMPLEX_ELEMENTARY_FUNCTIONS
B Rationale
    B.1 Introduction and motivation
    B.2 History
    B.3 Packaging considerations
    B.4 Choice of functions
    B.5 Handling of exceptions
    B.6 Matters of style
    B.7 Method of expressing accuracy requirements
C Bibliography
Figures
B.1 Result rectangle for a subprogram satisfying a maximum
    relative component error requirement
B.2 Result rectangle for a subprogram satisfying a maximum
    relative box error requirement
B.3 Result circle for a subprogram satisfying a maximum
    relative vector error requirement